About half of parents of children with ASD notice their child's unusual behaviors by age 18 months, and about four-fifths notice by age 24 months. According to an article, failure to meet any of the following milestones "is an absolute indication to proceed with further evaluations. Delay in referral for such testing may delay early diagnosis and treatment and affect the long-term outcome".

- No babbling by 12 months.

- No gesturing (pointing, waving, etc.) by 12 months.

- No single words by 16 months.

- No two-word (spontaneous, not just echolalic) phrases by 24 months.

- Any loss of any language or social skills, at any age.

The United States Preventive Services Task Force in 2016 found it was unclear if screening was beneficial or harmful among children in whom there is no concerns. The Japanese practice is to screen all children for ASD at 18 and 24 months, using autism-specific formal screening tests. In contrast, in the UK, children whose families or doctors recognize possible signs of autism are screened. It is not known which approach is more effective. Screening tools include the Modified Checklist for Autism in Toddlers (M-CHAT), the Early Screening of Autistic Traits Questionnaire, and the First Year Inventory; initial data on M-CHAT and its predecessor, the Checklist for Autism in Toddlers (CHAT), on children aged 18–30 months suggests that it is best used in a clinical setting and that it has low sensitivity (many false-negatives) but good specificity (few false-positives). It may be more accurate to precede these tests with a broadband screener that does not distinguish ASD from other developmental disorders. Screening tools designed for one culture's norms for behaviors like eye contact may be inappropriate for a different culture. Although genetic screening for autism is generally still impractical, it can be considered in some cases, such as children with neurological symptoms and dysmorphic features.

While infection with rubella during pregnancy causes fewer than 1% of cases of autism, vaccination against rubella can prevent many of those cases.

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.

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.

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.

There is a division among doctors on the use of the term PDD. Many use the term PDD as a short way of saying PDD-NOS. Others use the general category because the term PDD actually refers to a category of disorders and is not a diagnostic label.

PDD is not itself a diagnosis, while PDD-NOS is a diagnosis. To further complicate the issue, PDD-NOS can also be referred to as "atypical personality development", "atypical PDD", or "atypical Autism".

Because of the "NOS", which means "not otherwise specified", it is hard to describe what PDD-NOS is, other than its being an autism spectrum disorder (ASD). Some people diagnosed with PDD-NOS are close to having Asperger syndrome, but do not quite fit. Others have near full-fledged autism, but without some of its symptoms. The psychology field is considering creating several subclasses within PDD-NOS.

The pervasive developmental disorders are:

- Pervasive developmental disorder not otherwise specified (PDD-NOS), which includes atypical autism, and is the most common (47% of diagnoses);

- Autism, the best-known;

- Asperger syndrome (9% of autism diagnoses);

- Rett syndrome; and

- Childhood disintegrative disorder (CDD).

The first three of these disorders are commonly called the autism spectrum disorders; the last two disorders are much rarer, and are sometimes placed in the autism spectrum and sometimes not.

In May 2013, the "Diagnostic and Statistical Manual-Fifth Edition" ("DSM-5") was released, updating the classification for pervasive developmental disorders. The grouping of disorders, including PDD-NOS, Autism, Asperger Syndrome, Rett Syndrome, and CDD, has been removed and replaced with the general term of Autism Spectrum Disorders. The American Psychiatric Association has concluded that using the general diagnosis of ASD supports more accurate diagnoses. The combination of these disorders was also fueled by the standpoint that Autism is characterized by common symptoms and should therefore bear a single diagnostic term. In order to distinguish between the different disorders, the DSM-5 employs severity levels. The severity levels take into account required support, restricted interests and repetitive behaviors, and deficits in social communication.

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%.

PDD-NOS is an old diagnostic category. It is no longer included as an option for an Autism Spectrum Disorder and is not part of the DSM-5, but is included in the ICD-10.

The diagnosis of a pervasive developmental disorder not otherwise specified is given to individuals with difficulties in the areas of social interaction, communication, and/or stereotyped behavior patterns or interests, but who do not meet the full DSM-IV criteria for autism or another PDD. This does not necessarily mean that PDD-NOS is a milder disability than the other PDDs. It only means that individuals who receive this diagnosis do not meet the diagnostic criteria of the other PDDs, but that there is still a pervasive developmental disorder that affects the individual in the areas of communication, socialization and behavior.

As for the other pervasive developmental disorders, diagnosis of PDD-NOS requires the involvement of a team of specialists. The individual needs to undergo a full diagnostic evaluation, including a thorough medical, social, adaptive, motor skills and communication history. Other parts of an assessment can be behavioral rating scales, direct behavioral observations, psychological assessment, educational assessment, communication assessment, and occupational assessment.

Description of PDD-NOS merely as a "subthreshold" category without a more specific case definition poses methodological problems for research regarding the relatively heterogeneous group of people who receive this diagnosis. However, it appears that children with PDD-NOS show fewer intellectual deficits than autistic children, and that they may come to professional attention at a later age.

Studies suggest that persons with PDD-NOS belong to one of three very different subgroups:

- A high-functioning group (around 25 percent) whose symptoms largely overlap with that of Asperger syndrome, but who differ in terms of having a lag in language development and/or mild cognitive impairment. (The criteria for Asperger syndrome excludes a speech delay or a cognitive impairment.)

- A group (around 25 percent) whose symptoms more closely resemble those of autism spectrum disorder, but do not fully meet all its diagnostic signs and symptoms.

- The biggest group (around 50 percent) consists of those who meet all the diagnostic criteria for autism spectrum disorder, but whose stereotypical and repetitive behaviors are noticeably mild.

The first diagnosed case of ASD was published in 1943 by American psychiatrist Leo Kanner. There is a wide range of cases and severity to ASD so it is very hard to detect the first signs of ASD. A diagnosis of ASD can be made accurately before the child is 3 years old, but the diagnosis of ASD is not commonly confirmed until the child is somewhat older. The age of diagnosis can range from 9 months to 14 years, and the mean age is 4 years old in the USA. On average each case of ASD is tested at three different diagnostic centers before confirmed. Early diagnosis of the disorder can diminish familial stress, speed up referral to special educational programs and influence family planning.

Cytogenetic analysis for fragile X syndrome was first available in the late 1970s when diagnosis of the syndrome and carrier status could be determined by culturing cells in a folate deficient medium and then assessing for "fragile sites" (discontinuity of staining in the region of the trinucleotide repeat) on the long arm of the X chromosome. This technique proved unreliable, however, as the fragile site was often seen in less than 40% of an individual's cells. This was not as much of a problem in males, but in female carriers, where the fragile site could generally only be seen in 10% of cells, the mutation often could not be visualised.

Since the 1990s, more sensitive molecular techniques have been used to determine carrier status. The fragile X abnormality is now directly determined by analysis of the number of CGG repeats using polymerase chain reaction (PCR) and methylation status using Southern blot analysis. By determining the number of CGG repeats on the X chromosome, this method allows for more accurate assessment of risk for premutation carriers in terms of their own risk of fragile X associated syndromes, as well as their risk of having affected children. Because this method only tests for expansion of the CGG repeat, individuals with FXS due to missense mutations or deletions involving "FMR1" will not be diagnosed using this test and should therefore undergo sequencing of the FMR1 gene if there is clinical suspicion of FXS.

Prenatal testing with chorionic villus sampling or amniocentesis allows diagnosis of FMR1 mutation while the fetus is in utero and appears to be reliable.

Early diagnosis of fragile X syndrome or carrier status is important for providing early intervention in children or fetuses with the syndrome, and allowing genetic counselling with regards to the potential for a couple's future children to be affected. Most parents notice delays in speech and language skills, difficulties in social and emotional domains as well as sensitivity levels in certain situations with their children.

Developmental disorders comprise a group of psychiatric conditions originating in childhood that involve serious impairment in different areas. There are several ways of using this term. The most narrow concept is used in the category "Specific Disorders of Psychological Development" in the ICD-10. These disorders comprise language disorders, learning disorders, motor disorders and autism spectrum disorders. In broader definitions ADHD is included, and the term used is neurodevelopmental disorders. Yet others include antisocial behavior and schizophrenia that begins in childhood and continues through life. However, these two latter conditions are not as stable as the other developmental disorders, and there is not the same evidence of a shared genetic liability.

Developmental disorders are present from early life. They usually improve as the child grows older, but they also entail impairments that continue through adult life. There is a strong genetic component, and more males are afflicted than females.

There are no current treatments or cures for the underlying defects of FXS. Management of FXS may include speech therapy, behavioral therapy, sensory integration occupational therapy, special education, or individualised educational plans, and, when necessary, treatment of physical abnormalities. Persons with fragile X syndrome in their family histories are advised to seek genetic counseling to assess the likelihood of having children who are affected, and how severe any impairments may be in affected descendants.

Pathological Demand Avoidance is not recognised by the DSM-5 or ICD-10, the two main classification systems for mental disorder.

To be recognized a sufficient amount of consensus and clinical history needs to be present, and as a newly proposed condition PDA had not met the standard of evidence required at the time of recent revisions. In April 2014 the UK Minister of State for Care and Support Norman Lamb stated that the Department of Health, "In the course of the development of the National Institute for Health and Care Excellence (NICE) clinical guideline on the treatment of autism in children and young people (CG128), the developers looked at differential diagnoses for autism. In this, they did consider PDA, identifying it as a particular subgroup of autism that could also be described as oppositional defiant disorder (ODD). The guidance recommends that consideration should be given to differential diagnoses for autism (including ODD) and whether specific assessments are needed to help interpret the autism history and observations. However, due to the lack of evidence and the fact that the syndrome is not recognised within the DSM or ICD classifications, NICE was unable to develop specific recommendations on the assessment and treatment of PDA."

So the National Institute for Health and Care Excellence (which provides guidelines on best practice for UK clinicians) makes no mention of PDA in its guidelines for diagnosis of autism either in children or adults.

It is estimated that 25 to 50% of children diagnosed with Autism Spectrum Disorder (ASD) never develop spoken language beyond a few words or utterances. Despite the growing field of research on ASD, there is not much information available pertaining to individuals with autism who never develop functional language; that, in fact, individuals with nonverbal autism are considered to be underrepresented in all of autism research. Because of the limited research on nonverbal autism, there are not many validated measurements appropriate for this population. For example, while they may be appropriate for younger children, they lack the validity for grade-school aged children and adolescents and have continued to be a roadblock for nonverbal autism research. Often in autism research, individuals with nonverbal autism are sub-grouped with LFA, categorized by learning at most one word or having minimal verbal language.

Most of the existing body of research in nonverbal autism focuses on early interventions that predict successful language outcomes. Research suggests that most spoken language is inherited before the age of five, and the likelihood of acquiring functional language in the future past this age is minimal, that early language development is crucial to educational achievement, employment, independence during adulthood, and social relationships.

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.

The diagnosis of CdLS is primarily a clinical one, based on medical signs that are evident in a medical history, physical examination, and laboratory tests. Since 2006, testing for NIPBL and SMC1A has been available through the University of Chicago. This is best accomplished through a referral to a genetics specialist or clinic.

CdLS is thought to be underdiagnosed and frequently misdiagnosed.

The extra chromosome in people with idic(15) can be easily detected through chromosome analysis (karyotyping). Additional tests are usually required. FISH (Fluorescent in situ hybridization) is used to confirm the diagnosis by distinguishing idic(15) from other supernumerary marker chromosomes. Array CGH can be used to determine the gene content and magnitude of copy number variation so that the clinical picture can be foreseen.

Interstitial duplications of chromosome 15 can be more difficult to detect on a routine chromosome analysis but are clearly identifiable using a 15q FISH study. Families should always discuss the results of chromosome and FISH studies with a genetic counselor or other genetics professionals to ensure accurate interpretation.

In general, idic(15) occurs de novo but the parents must be karyotyped to make sure it is not inherited, mostly because this will affect the course of genetic counseling given to the family. If the abnormality is found prenatally and one of the parents harbour the marker, the child has a chance of not carrying the mutation. Further tests should however be done to prove the marker has not been rearranged while being inherited. This information is also necessary for counseling of future pregnancies. Each family is unique and should therefore be handled individually.

For nonverbal grade school children and adolescents with autism, communication systems and interventions have been implemented to enhance language and communication outcomes. Speech-generated devices, such as iPads, use visual displays for children who lack verbal language, giving them the task of selecting icons indicating a request or need. For adolescents with nonverbal autism, interventions can condition them to learn more advanced operations on speech-generated devices that require more steps (i.e. turning on device, scrolling through pages), which would allow them to enhance their communicative abilities independently.

The Picture Exchange System (PECS) is an alternative form of spontaneous communication for children with autism in which an individual selects a picture indicating a request. PECS can be utilized in educational settings and at the child’s home. Longitudinal studies suggest PECS can have long-term positive outcomes for school-aged children with nonverbal autism, specifically their social-communicative skills, such as higher frequencies of joint attention and initiation, and duration of cooperative play, which are all important roles in improving language outcomes.

It has also been suggested that a significant stage in acquiring verbal language is learning how to identify and reproduce syllables of words. One study found that nonverbal and minimally verbal children with autism are capable of enhancing their oral production and vocalizing written words by isolating each syllable of a word one at a time. The process of breaking down a syllable at a time and having it visually displayed and audibly available to the child can prompt him or her to imitate and create nonrandom and meaningful utterances.

Most of these studies contain small sample sizes and were pilot studies, making additional research significant to assess whether these findings can be generalized to all age groups of the same population. Furthermore, most studies on nonverbal autism speech-generated device communication were based on more basic skills, such as naming pictures and making requests for stimuli, while studies in advanced communication (i.e. asking "how are you?") is limited.

Pragmatic language impairment (PLI), or social (pragmatic) communication disorder (SCD), is an impairment in understanding pragmatic aspects of language. This type of impairment was previously called semantic-pragmatic disorder (SPD). People with these impairments have special challenges with the semantic aspect of language (the meaning of what is being said) and the pragmatics of language (using language appropriately in social situations). It is assumed that those with autism have difficulty with "the meaning of what is being said" due to different ways of responding to social situations.

PLI is now a diagnosis in DSM-5, and is called social (pragmatic) communication disorder. Communication problems are also part of the autism spectrum disorders (ASD); however, the latter also show a restricted pattern of behavior, according to behavioral psychology. The diagnosis SCD can only be given if ASD has been ruled out.

Diagnosing Jacobsen Syndrome can be difficult in some cases because it is a rare chromosomal disorder. There are a variety of tests that can be carried out like karyotype, cardiac echocardiogram, a renal sonogram, a platelet count, blood count, a brain imaging study. Genetic testing can be carried out for diagnosis. In which chromosomes are stained to give a barcode like appearance and studied under the microscope which reveals the broken and deleted genes. It can also be diagnosed early in the prenatal stage if there are any abnormalities seen in the ultrasound. A simple assessment of the symptoms can be done to diagnose the Syndrome. A thorough physical examination could be carried out to assess the symptoms.

Pathological demand avoidance (PDA) is a proposed subtype of autism characterized by an avoidance of demand-framed requests by an individual. It was proposed in 1980 by the UK child psychologist Elizabeth Newson. The Elizabeth Newson Centre in Nottingham, England carries out assessments for the NHS, local authorities and private patients around autism spectrum disorder, which include, but are not exclusively PDA.

PDA behaviours are consistent with autism, but have differences from other autism subtypes diagnoses. It is not recognised by either the DSM-5 or the .

In 1983, Rapin and Allen suggested the term "semantic pragmatic disorder" to describe the communicative behavior of children who presented traits such as pathological talkativeness, deficient access to vocabulary and discourse comprehension, atypical choice of terms and inappropriate conversational skills. They referred to a group of children who presented with mild autistic features and specific semantic pragmatic language problems. More recently, the term "pragmatic language impairment" (PLI) has been proposed.

Rapin and Allen's definition has been expanded and refined by therapists who include communication disorders that involve difficulty in understanding the meaning of words, grammar, syntax, prosody, eye gaze, body language, gestures, or social context. While autistic children exhibit pragmatic language impairment, this type of communication disorder can also be found in individuals with other types of disorders including auditory processing disorders, neuropathies, encephalopathies and certain genetic disorders.