Developmental disabilities can be initially suspected when a child does not reach expected child development stages. Subsequently, a differential diagnosis may be used to diagnose an underlying disease, which may include a physical examination and genetic tests.

The degree of disability can be quantified by assigning a "developmental age" to a person, which is age of the group into which test scores place the person. This, in turn, can be used to calculate a "" (DQ) as follows:

formula_1

The first English-language IQ test, the Stanford–Binet Intelligence Scales, was adapted from a test battery designed for school placement by Alfred Binet in France. Lewis Terman adapted Binet's test and promoted it as a test measuring "general intelligence." Terman's test was the first widely used mental test to report scores in "intelligence quotient" form ("mental age" divided by chronological age, multiplied by 100). Current tests are scored in "deviation IQ" form, with a performance level by a test-taker two standard deviations below the median score for the test-taker's age group defined as IQ 70. Until the most recent revision of diagnostic standards, an IQ of 70 or below was a primary factor for intellectual disability diagnosis, and IQ scores were used to categorize degrees of intellectual disability.

Since current diagnosis of intellectual disability is not based on IQ scores alone, but must also take into consideration a person's adaptive functioning, the diagnosis is not made rigidly. It encompasses intellectual scores, adaptive functioning scores from an adaptive behavior rating scale based on descriptions of known abilities provided by someone familiar with the person, and also the observations of the assessment examiner who is able to find out directly from the person what he or she can understand, communicate, and such like. IQ assessment must be based on a current test. This enables diagnosis to avoid the pitfall of the Flynn effect, which is a consequence of changes in population IQ test performance changing IQ test norms over time.

Adaptive behavior, or adaptive functioning, refers to the skills needed to live independently (or at the minimally acceptable level for age). To assess adaptive behavior, professionals compare the functional abilities of a child to those of other children of similar age. To measure adaptive behavior, professionals use structured interviews, with which they systematically elicit information about persons' functioning in the community from people who know them well. There are many adaptive behavior scales, and accurate assessment of the quality of someone's adaptive behavior requires clinical judgment as well. Certain skills are important to adaptive behavior, such as:

- Daily living skills, such as getting dressed, using the bathroom, and feeding oneself

- Communication skills, such as understanding what is said and being able to answer

- Social skills with peers, family members, spouses, adults, and others

There are many physical health factors associated with developmental disabilities. For some specific syndromes and diagnoses, these are inherent, such as poor heart function in people with Down syndrome. People with severe communication difficulties find it difficult to articulate their health needs, and without adequate support and education might not recognize ill health. Epilepsy, sensory problems (such as poor vision and hearing), obesity and poor dental health are over-represented in this population. Life expectancy among people with developmental disabilities as a group is estimated at 20 years below average, although this is improving with advancements in adaptive and medical technologies, and as people are leading healthier, more fulfilling lives, and some conditions (such as Freeman-Sheldon syndrome) do not impact life expectancy.

Traditionally, genetic abnormalities in neurodevelopmental disorders were detected using karyotype analysis, which found 5% of relevant disorders. , chromosomal microarray analysis (CMA) has replaced karyotyping, because of its greater diagnostic yield in about 20% of cases, detecting smaller chromosome abnormalities. It is the first line genomic test.

New descriptions include the term Copy-number variants (CNVs), which are losses or gains of chromosomal regions greater than 1 kb in length. CNVs are mentioned with the chromosomal band(s) they involve and their genome sequence coordinates. CNVs can be nonrecurrent and recurrent.

With CMA costs of testing have increased from 800 US$ to 1500$. Guidelines from the American College of Medical Genetics and Genomics and the American Academy of Pediatrics recommend CMA as standard of care in the US.

Neurodevelopmental disorders are in their multitude associated with widely varying degrees of difficulty, depending on which there are different degrees of mental, emotional, physical, and economic consequences for individuals, and in turn families, groups and society.

ASD can be detected as early as 18 months or even younger in some cases. A reliable diagnosis can usually be made by the age of two years. The diverse expressions of ASD symptoms pose diagnostic challenges to clinicians. Individuals with an ASD may present at various times of development (e.g., toddler, child, or adolescent), and symptom expression may vary over the course of development. Furthermore, clinicians must differentiate among pervasive developmental disorders, and may also consider similar conditions, including intellectual disability not associated with a pervasive developmental disorder, specific language disorders, ADHD, anxiety, and psychotic disorders.

Considering the unique challenges in diagnosing ASD, specific practice parameters for its assessment have been published by the American Academy of Neurology, the American Academy of Child and Adolescent Psychiatry, and a consensus panel with representation from various professional societies. The practice parameters outlined by these societies include an initial screening of children by general practitioners (i.e., "Level 1 screening") and for children who fail the initial screening, a comprehensive diagnostic assessment by experienced clinicians (i.e. "Level 2 evaluation"). Furthermore, it has been suggested that assessments of children with suspected ASD be evaluated within a developmental framework, include multiple informants (e.g., parents and teachers) from diverse contexts (e.g., home and school), and employ a multidisciplinary team of professionals (e.g., clinical psychologists, neuropsychologists, and psychiatrists).

After a child shows initial evidence of ASD tendencies, psychologists administer various psychological assessment tools to assess for ASD. Among these measurements, the Autism Diagnostic Interview-Revised (ADI-R) and the Autism Diagnostic Observation Schedule (ADOS) are considered the "gold standards" for assessing autistic children. The ADI-R is a semi-structured parent interview that probes for symptoms of autism by evaluating a child's current behavior and developmental history. The ADOS is a semistructured interactive evaluation of ASD symptoms that is used to measure social and communication abilities by eliciting several opportunities (or "presses") for spontaneous behaviors (e.g., eye contact) in standardized context. Various other questionnaires (e.g., The Childhood Autism Rating Scale, Autism Treatment Evaluation Checklist) and tests of cognitive functioning (e.g., The Peabody Picture Vocabulary Test) are typically included in an ASD assessment battery.

In the UK, there is some diagnostic use of the Diagnostic Interview for Social and Communication Disorders (DISCO) was which was developed for use at The Centre for Social and Communication Disorders, by Lorna Wing and Judith Gould, as both a clinical and a research instrument for use with children and adults of any age. The DISCO is designed to elicit a picture of the whole person through the story of their development and behaviour. In clinical work, the primary purpose is to facilitate understanding of the pattern over time of the specific skills and impairments that underlie the overt behaviour. If no information is available, the clinician has to obtain as much information as possible concerning the details of current skills and pattern of behaviour of the person. This type of dimensional approach to clinical description is useful for prescribing treatment.

Parents of children with Asperger syndrome can typically trace differences in their children's development to as early as 30 months of age. Developmental screening during a routine check-up by a general practitioner or pediatrician may identify signs that warrant further investigation. The United States Preventive Services Task Force in 2016 found it was unclear if screening was beneficial or harmful among children in whom there are no concerns.

The diagnosis of AS is complicated by the use of several different screening instruments, including the Asperger Syndrome Diagnostic Scale (ASDS), Autism Spectrum Screening Questionnaire (ASSQ), Childhood Autism Spectrum Test (CAST) (previously called the Childhood Asperger Syndrome Test), Gilliam Asperger's disorder scale (GADS), Krug Asperger's Disorder Index (KADI), and the Autism-spectrum quotient (AQ; with versions for children, adolescents and adults). None have been shown to reliably differentiate between AS and other ASDs.

Standard diagnostic criteria require impairment in social interaction and repetitive and stereotyped patterns of behavior, activities and interests, without significant delay in language or cognitive development. Unlike the international standard, the DSM-IV-TR criteria also required significant impairment in day-to-day functioning; DSM-5 eliminated AS as a separate diagnosis in 2013, and folded it into the umbrella of autism spectrum disorders. Other sets of diagnostic criteria have been proposed by Szatmari "et al." and by Gillberg and Gillberg.

Diagnosis is most commonly made between the ages of four and eleven. A comprehensive assessment involves a multidisciplinary team that observes across multiple settings, and includes neurological and genetic assessment as well as tests for cognition, psychomotor function, verbal and nonverbal strengths and weaknesses, style of learning, and skills for independent living. The "gold standard" in diagnosing ASDs combines clinical judgment with the Autism Diagnostic Interview-Revised (ADI-R)—a semistructured parent interview—and the Autism Diagnostic Observation Schedule (ADOS)—a conversation and play-based interview with the child. Delayed or mistaken diagnosis can be traumatic for individuals and families; for example, misdiagnosis can lead to medications that worsen behavior.

Underdiagnosis and overdiagnosis may be problems. The cost and difficulty of screening and assessment can delay diagnosis. Conversely, the increasing popularity of drug treatment options and the expansion of benefits has motivated providers to overdiagnose ASD. There are indications AS has been diagnosed more frequently in recent years, partly as a residual diagnosis for children of normal intelligence who are not autistic but have social difficulties.

There are questions about the external validity of the AS diagnosis. That is, it is unclear whether there is a practical benefit in distinguishing AS from HFA and from PDD-NOS; the same child can receive different diagnoses depending on the screening tool. The debate about distinguishing AS from HFA is partly due to a tautological dilemma where disorders are defined based on severity of impairment, so that studies that appear to confirm differences based on severity are to be expected.

Many normed assessments can be used in evaluating skills in the primary academic domains: reading, including word recognition, fluency, and comprehension; mathematics, including computation and problem solving; and written expression, including handwriting, spelling and composition.

The most commonly used comprehensive achievement tests include the Woodcock-Johnson IV (WJ IV), Wechsler Individual Achievement Test II (WIAT II), the Wide Range Achievement Test III (WRAT III), and the Stanford Achievement Test–10th edition. These tests include measures of many academic domains that are reliable in identifying areas of difficulty.

In the reading domain, there are also specialized tests that can be used to obtain details about specific reading deficits. Assessments that measure multiple domains of reading include Gray's Diagnostic Reading Tests–2nd edition (GDRT II) and the Stanford Diagnostic Reading Assessment. Assessments that measure reading subskills include the Gray Oral Reading Test IV – Fourth Edition (GORT IV), Gray Silent Reading Test, Comprehensive Test of Phonological Processing (CTOPP), Tests of Oral Reading and Comprehension Skills (TORCS), Test of Reading Comprehension 3 (TORC-3), Test of Word Reading Efficiency (TOWRE), and the Test of Reading Fluency. A more comprehensive list of reading assessments may be obtained from the Southwest Educational Development Laboratory.

The purpose of assessment is to determine what is needed for intervention, which also requires consideration of contextual variables and whether there are comorbid disorders that must also be identified and treated, such as behavioral issues or language delays. These contextual variables are often assessed using parent and teacher questionnaire forms that rate the students' behaviors and compares them to standardized norms.

However, caution should be made when suspecting the person with a learning disability may also have dementia, especially as people with Down's syndrome may have the neuroanatomical profile but not the associated clinical signs and symptoms. Examination can be carried out of executive functioning as well as social and cognitive abilities but may need adaptation of standardized tests to take account of special needs.

For many adopted or adults and children in foster care, records or other reliable sources may not be available for review. Reporting alcohol use during pregnancy can also be stigmatizing to birth mothers, especially if alcohol use is ongoing. In these cases, all diagnostic systems use an unknown prenatal alcohol exposure designation. A diagnosis of FAS is still possible with an unknown exposure level if other key features of FASD are present at clinical levels.

About 92% of pregnancies in Europe with a diagnosis of Down syndrome are terminated. In the United States, termination rates are around 67%, but this rate varied from 61% to 93% among different populations evaluated. When nonpregnant people are asked if they would have a termination if their fetus tested positive, 23–33% said yes, when high-risk pregnant women were asked, 46–86% said yes, and when women who screened positive are asked, 89–97% say yes.

Confirmed absence of exposure would apply to planned pregnancies in which no alcohol was used or pregnancies of women who do not use alcohol or report no use during the pregnancy. This designation is relatively rare, as most people presenting for an FASD evaluation are at least "suspected" to have had a prenatal alcohol exposure due to presence of other key features of FASD.

Autism spectrum disorders tend to be highly comorbid with other disorders. Comorbidity may increase with age and may worsen the course of youth with ASDs and make intervention/treatment more difficult. Distinguishing between ASDs and other diagnoses can be challenging, because the traits of ASDs often overlap with symptoms of other disorders, and the characteristics of ASDs make traditional diagnostic procedures difficult.

The most common medical condition occurring in individuals with autism spectrum disorders is seizure disorder or epilepsy, which occurs in 11-39% of individuals with ASD. Tuberous sclerosis, a medical condition in which non-malignant tumors grow in the brain and on other vital organs, occurs in 1-4% of individuals with ASDs.

Intellectual disabilities are some of the most common comorbid disorders with ASDs. Recent estimates suggest that 40-69% of individuals with ASD have some degree of an intellectual disability, more likely to be severe for females. A number of genetic syndromes causing intellectual disability may also be comorbid with ASD, including fragile X syndrome, Down syndrome, Prader-Willi and Angelman syndromes, and Williams syndrome.

Learning disabilities are also highly comorbid in individuals with an ASD. Approximately 25-75% of individuals with an ASD also have some degree of a learning disability.

Various anxiety disorders tend to co-occur with autism spectrum disorders, with overall comorbidity rates of 7-84%. Rates of comorbid depression in individuals with an ASD range from 4–58%. The relationship between ASD and schizophrenia remains a controversial subject under continued investigation, and recent meta-analyses have examined genetic, environmental, infectious, and immune risk factors that may be shared between the two conditions.

Deficits in ASD are often linked to behavior problems, such as difficulties following directions, being cooperative, and doing things on other people's terms. Symptoms similar to those of attention deficit hyperactivity disorder (ADHD) can be part of an ASD diagnosis.

Sensory processing disorder is also comorbid with ASD, with comorbidity rates of 42–88%.

Learning disability is a classification that includes several areas of functioning in which a person has difficulty learning in a typical manner, usually caused by an unknown factor or factors. Given the "difficulty learning in a typical manner", this does not exclude the ability to learn in a different manner. Therefore, some people can be more accurately described as having a "Learning Difference", thus avoiding any misconception of being disabled with a lack of ability to learn and possible negative stereotyping.

In the UK, the term "learning disability" generally refers to an intellectual disability, while difficulties such as dyslexia and dyspraxia are usually referred to as "learning difficulties".

While "learning disability, learning disorder" and "learning difficulty" are often used interchangeably, they differ in many ways. Disorder refers to significant learning problems in an academic area. These problems, however, are not enough to warrant an official diagnosis. Learning disability, on the other hand, is an official clinical diagnosis, whereby the individual meets certain criteria, as determined by a professional (psychologist, pediatrician, etc.). The difference is in degree, frequency, and intensity of reported symptoms and problems, and thus the two should not be confused. When the term "learning disorder" is used, it describes a group of disorders characterized by inadequate development of specific academic, language, and speech skills. Types of learning disorders include reading (dyslexia), mathematics (dyscalculia) and writing (dysgraphia).

The unknown factor is the disorder that affects the brain's ability to receive and process information. This disorder can make it problematic for a person to learn as quickly or in the same way as someone who is not affected by a learning disability. People with a learning disability have trouble performing specific types of skills or completing tasks if left to figure things out by themselves or if taught in conventional ways.

Individuals with learning disabilities can face unique challenges that are often pervasive throughout the lifespan. Depending on the type and severity of the disability, interventions, and current technologies may be used to help the individual learn strategies that will foster future success. Some interventions can be quite simplistic, while others are intricate and complex. Current technologies may require student training to be effective classroom supports. Teachers, parents, and schools can create plans together that tailor intervention and accommodations to aid the individuals in successfully becoming independent learners. School psychologists and other qualified professionals quite often help design the intervention and coordinate the execution of the intervention with teachers and parents.

The diagnosis can often be suspected based on the child's physical appearance at birth. An analysis of the child's chromosomes is needed to confirm the diagnosis, and to determine if a translocation is present, as this may help determine the risk of the child's parents having further children with Down syndrome. Parents generally wish to know the possible diagnosis once it is suspected and do not wish pity.

Assessment will usually include an interview with the child’s caregiver, observation of the child in an unstructured setting, a hearing test, and standardized tests of language. There is a wide range of language assessments in English. Some are restricted for use by experts in speech-language pathology: speech and language therapists (SaLTs/SLTs) in the UK, speech-language pathologists (SLPs) in the US and Australia. A commonly used test battery for diagnosis of DLD is the Clinical Evaluation of Language Fundamentals (CELF).

Assessments that can be completed by a parent or teacher can be useful to identify children who may require more in-depth evaluation. The Children’s Communication Checklist (CCC–2) is a parent questionnaire suitable for assessing everyday use of language in children aged 4 years and above who can speak in sentences.

Informal assessments, such as language samples, are often used by speech-language therapists/pathologists to complement formal testing and give an indication of the child's language in a more naturalistic context. A language sample may be of a conversation or narrative retell. In a narrative language sample, an adult may tell the child a story using a wordless picture book (e.g. Frog Where Are You?, Mayer, 1969), then ask the child to use the pictures and tell the story back. Language samples can be transcribed using computer software such as the Systematic Analysis of Language Software, and then analyzed for a range of features: e.g., the grammatical complexity of the child's utterances, whether the child introduces characters to their story or jumps right in, whether the events follow a logical order, and whether the narrative includes a main idea or theme and supporting details.

At its most basic level, dyscalculia is a learning disability affecting the normal development of arithmetic skills.

A consensus has not yet been reached on appropriate diagnostic criteria for dyscalculia. Mathematics is a specific domain that is complex (i.e. includes many different processes, such as arithmetic, algebra, word problems, geometry, etc.) and cumulative (i.e. the processes build on each other such that mastery of an advanced skill requires mastery of many basic skills). Thus dyscalculia can be diagnosed using different criteria, and frequently is; this variety in diagnostic criteria leads to variability in identified samples, and thus variability in research findings regarding dyscalculia.

Other than using achievement tests as diagnostic criteria, researchers often rely on domain-specific tests (i.e. tests of working memory, executive function, inhibition, intelligence, etc.) and teacher evaluations to create a more comprehensive diagnosis. Alternatively, fMRI research has shown that the brains of the neurotypical children can be reliably distinguished from the brains of the dyscalculic children based on the activation in the prefrontal cortex. However, due to the cost and time limitations associated with brain and neural research, these methods will likely not be incorporated into diagnostic criteria despite their effectiveness.

There are tests that can indicate with high probability whether a person is a dyslexic. If diagnostic testing indicates that a person may be dyslexic, such tests are often followed up with a full diagnostic assessment to determine the extent and nature of the disorder. Tests can be administered by a teacher or computer. Some test results indicate how to carry out teaching strategies.

There are a variety of medical conditions affecting cognitive ability. This is a broad concept encompassing various intellectual or cognitive deficits, including intellectual disability, deficits too mild to properly qualify as intellectual disability, various specific conditions (such as specific learning disability), and problems acquired later in life through acquired brain injuries or neurodegenerative diseases like dementia. These disabilities may appear at any age.

Dyslexic children require special instruction for word analysis and spelling from an early age. While there are fonts that may help people with dyslexia better understand writing, this might simply be due to the added spacing between words. The prognosis, generally speaking, is positive for individuals who are identified in childhood and receive support from friends and family.

To date, very few interventions have been developed specifically for individuals with dyscalculia. Concrete manipulation activities have been used for decades to train basic number concepts for remediation purposes. This method facilitates the intrinsic relationship between a goal, the learner’s action, and the informational feedback on the action. A one-to-one tutoring paradigm designed by Lynn Fuchs and colleagues which teaches concepts in arithmetic, number concepts, counting, and number families using games, flash cards, and manipulables has proven successful in children with generalized math learning difficulties, but intervention has yet to be tested specifically on children with dyscalculia. These methods require specially trained teachers working directly with small groups or individual students. As such, instruction time in the classroom is necessarily limited. For this reason, several research groups have developed computer adaptive training programs designed to target deficits unique to dyscalculic individuals.

Software intended to remediate dyscalculia has been developed. While computer adaptive training programs are modeled after one-to-one type interventions, they provide several advantages. Most notably, individuals are able to practice more with a digital intervention than is typically possible with a class or teacher. As with one-to-one interventions, several digital interventions have also proven successful in children with generalized math learning difficulties. Räsänen and colleagues have found that games such as The Number Race and Graphogame-math can improve performance on number comparison tasks in children with generalized math learning difficulties.

Several digital interventions have been developed for dyscalculics specifically. Each attempts to target basic processes that are associated with maths difficulties. Rescue Calcularis was one early computerized intervention that sought to improve the integrity of and access to the mental number line. Other digital interventions for dyscalculia adapt games, flash cards, and manipulables to function through technology.

While each intervention claims to improve basic numerosity skills, the authors of these interventions do admit that repetition and practice effects may be a factor involved in reported performance gains. An additional criticism is that these digital interventions lack the option to manipulate numerical quantities. While the previous two games provide the correct answer, the individual using the intervention cannot actively determine, through manipulation, what the correct answer should be. Butterworth and colleagues argued that games like The Number Bonds, which allows an individual to compare different sized rods, should be the direction that digital interventions move towards. Such games use manipulation activities to provide intrinsic motivation towards content guided by dyscalculia research. One of these serious games is Meister Cody – Talasia, an online training that includes the CODY Assessment – a diagnostic test for detecting dyscalculia. Based on these findings, Rescue Calcluaris was extended by adaptation algorithms and game forms allowing manipulation by the learners. It was found to improve addition, subtraction and number line tasks, and was made available as Dybuster Calcularis.

A study used transcranial direct current stimulation (TDCS) to the parietal lobe during numerical learning and demonstrated selective improvement of numerical abilities that was still present six months later in typically developing individuals. Improvement were achieved by applying anodal current to the right parietal lobe and cathodal current to the left parietal lobe and contrasting it with the reverse setup. When the same research group used tDCS in a training study with two dyscalculic individuals, the reverse setup (left anodal, right cathodal) demonstrated improvement of numerical abilities.

Intellectual disability in children can be caused by genetic or environmental factors. The individual could have a natural brain malformation or pre or postnatal damage done to the brain caused by drowning or a traumatic brain injury, for example. Nearly 30 to 50% of individuals with intellectual disability will never know the cause of their diagnosis even after thorough investigation.

Prenatal causes of intellectual disability include:

- Congenital infections such as cytomegalovirus, toxoplasmosis, herpes, syphilis, rubella and human immunodeficiency virus

- Prolonged maternal fever in the first trimester

- Exposure to anticonvulsants or alcohol

- Untreated maternal phenylketonuria (PKU)

- Complications of prematurity, especially in extremely low-birth-weight infants

- Postnatal exposure to lead

Single-gene disorders that result in intellectual disability include:

- Fragile X syndrome

- Neurofibromatosis

- Tuberous sclerosis

- Noonan's syndrome

- Cornelia de Lange's syndrome

These single-gene disorders are usually associated with atypical physical characteristics.

About 1/4 of individuals with intellectual disability have a detectable chromosomal abnormality. Others may have small amounts of deletion or duplication of chromosomes, which may go unnoticed and therefore, undetermined.

Late talker is a term used for exceptionally bright people who experience a delay in the development of speech. Commonalities include usually being boys, delayed speech development, highly educated parents, musically gifted families, puzzle-solving abilities, and lagging social development. Many high-achieving late talkers were notoriously strong willed and noncompliant as children. Late talkers can often be misdiagnosed early on as having severe ("low-functioning") autism spectrum disorder (a category known simply as "autism", prior to the DSM-5), and careful professional evaluation is necessary for differential diagnosis, according to Darold Treffert and other experts. One major difference between late talkers and low-functioning autistic children is that for late talkers, communication skills automatically reach a normal level and the child requires no further special treatment with regards to speech. Outlook for late talkers with or without intervention is generally favorable. However, late language emergence can also be an early or secondary sign of high-functioning autism spectrum disorder / Asperger syndrome, or other developmental disorders, such as attention deficit hyperactivity disorder, intellectual disability, learning disability, social communication disorder, or specific language impairment.

Einstein syndrome, a term coined by the economist Thomas Sowell, is also sometimes used to describe late talkers. The term is named after Albert Einstein (often said to have been a late talker, though with questionable evidence), whom Sowell used as the primary example of a late talker in his work. Sowell also included Edward Teller, Srinivasa Ramanujan, the mathematician Julia Robinson, Richard Feynman, and the pianists Clara Schumann and Arthur Rubinstein to be in the late talkers group. As a toddler, the scientist John Clive Ward showed similar behavioral traits to those described by Sowell, according to a brief sketch of his biography.

Sowell claimed late talkers are often inaccurately categorized as having an autism spectrum disorder (ASD), and that a small subset of late talkers are highly intelligent children with common characteristics concentrated in music, memory, math or the sciences. However, as reported by Simon Baron-Cohen, such characteristics are often found in high-functioning autism / Asperger syndrome.

DLD is defined purely in behavioural terms: there is no biological test. There are three points that need to be met for a diagnosis of DLD:

1. The child has language difficulties that create obstacles to communication or learning in everyday life,

2. The child's language problems are unlikely to resolve by five years of age, and

3. The problems are not associated with a known biomedical condition such as brain injury, neurodegenerative conditions, genetic conditions or chromosome disorders such as Down Syndrome, sensorineural hearing loss, or Autism Spectrum Disorder or Intellectual Disability.

For research and epidemiological purposes, specific cutoffs on language assessments have been used to document the first criterion. Tomblin et al. proposed the EpiSLI criterion, based on five composite scores representing performance in three domains of language (vocabulary, grammar, and narration) and two modalities (comprehension and production). Children scoring in the lowest 10% on two or more composite scores are identified as having language disorder.

The second criterion, persistence of language problems, can be difficult to judge in a young child, but longitudinal studies have shown that difficulties are less likely to resolve for children who have poor language comprehension, rather than difficulties confined to expressive language. In addition, children with isolated difficulties in just one of the areas noted under 'subtypes' tend to make better progress than those whose language is impaired in several areas.

The third criterion specifies that DLD is used for children whose language disorder is not part of another biomedical condition, such as a genetic syndrome, a sensorineural hearing loss, neurological disease, Autism Spectrum Disorder or Intellectual Disability – these were termed 'differentiating conditions' by the CATALISE panel. Language disorders occurring with these conditions need to be assessed and children offered appropriate intervention, but a terminological distinction is made so that these cases would be diagnosed as Language Disorder associated with ___, with the main diagnosis being specified: e.g. "Language Disorder associated with Autism Spectrum Disorder." The reasoning behind these diagnostic distinctions is discussed further by Bishop (2017).