The Düsseldorf score stratifies cases using four categories, giving one point for each; bone marrow blasts ≥5%, LDH >200U/L, haemoglobin ≤9g/dL and a platelet count ≤100,000/uL. A score of 0 indicates a low risk group' 1-2 indicates an intermediate risk group and 3-4 indicates a high risk group. The cumulative 2 year survival of scores 0, 1-2 and 3-4 is 91%, 52% and 9%; and risk of AML transformation is 0%, 19% and 54% respectively.

Evidence is conflicting on the prognostic significance of chloromas in patients with acute myeloid leukemia. In general, they are felt to augur a poorer prognosis, with a poorer response to treatment and worse survival; however, others have reported chloromas associate, as a biologic marker, with other poor prognostic factors, and therefore do not have independent prognostic significance.

The first clue to a diagnosis of AML is typically an abnormal result on a complete blood count. While an excess of abnormal white blood cells (leukocytosis) is a common finding with the leukemia, and leukemic blasts are sometimes seen, AML can also present with isolated decreases in platelets, red blood cells, or even with a low white blood cell count (leukopenia). While a presumptive diagnosis of AML can be made by examination of the peripheral blood smear when there are circulating leukemic blasts, a definitive diagnosis usually requires an adequate bone marrow aspiration and biopsy as well as ruling out pernicious anemia (Vitamin B12 deficiency), folic acid deficiency and copper deficiency.

Marrow or blood is examined under light microscopy, as well as flow cytometry, to diagnose the presence of leukemia, to differentiate AML from other types of leukemia (e.g. acute lymphoblastic leukemia - ALL), and to classify the subtype of disease. A sample of marrow or blood is typically also tested for chromosomal abnormalities by routine cytogenetics or fluorescent "in situ" hybridization. Genetic studies may also be performed to look for specific mutations in genes such as "FLT3", nucleophosmin, and "KIT", which may influence the outcome of the disease.

Cytochemical stains on blood and bone marrow smears are helpful in the distinction of AML from ALL, and in subclassification of AML. The combination of a myeloperoxidase or Sudan black stain and a nonspecific esterase stain will provide the desired information in most cases. The myeloperoxidase or Sudan black reactions are most useful in establishing the identity of AML and distinguishing it from ALL. The nonspecific esterase stain is used to identify a monocytic component in AMLs and to distinguish a poorly differentiated monoblastic leukemia from ALL.

The diagnosis and classification of AML can be challenging, and should be performed by a qualified hematopathologist or hematologist. In straightforward cases, the presence of certain morphologic features (such as Auer rods) or specific flow cytometry results can distinguish AML from other leukemias; however, in the absence of such features, diagnosis may be more difficult.

The two most commonly used classification schemata for AML are the older French-American-British (FAB) system and the newer World Health Organization (WHO) system. According to the widely used WHO criteria, the diagnosis of AML is established by demonstrating involvement of more than 20% of the blood and/or bone marrow by leukemic myeloblasts, except in the three best prognosis forms of acute myeloid leukemia with recurrent genetic abnormalities (t(8;21), inv(16), and t(15;17)) in which the presence of the genetic abnormality is diagnostic irrespective of blast percent. The French–American–British (FAB) classification is a bit more stringent, requiring a blast percentage of at least 30% in bone marrow (BM) or peripheral blood (PB) for the diagnosis of AML. AML must be carefully differentiated from "preleukemic" conditions such as myelodysplastic or myeloproliferative syndromes, which are treated differently.

Because acute promyelocytic leukemia (APL) has the highest curability and requires a unique form of treatment, it is important to quickly establish or exclude the diagnosis of this subtype of leukemia. Fluorescent "in situ" hybridization performed on blood or bone marrow is often used for this purpose, as it readily identifies the chromosomal translocation [t(15;17)(q22;q12);] that characterizes APL. There is also a need to molecularly detect the presence of PML/RARA fusion protein, which is an oncogenic product of that translocation.

Diagnosing ALL begins with a thorough medical history, physical examination, complete blood count, and blood smears. While many symptoms of ALL can be found in common illnesses, persistent or unexplained symptoms raise suspicion of cancer. Because many features on the medical history and exam are not specific to ALL, further testing is often needed. A large number of white blood cells and lymphoblasts in the circulating blood can be suspicious for ALL because they indicate a rapid production of lymphoid cells in the marrow. The higher these numbers typically points to a worse prognosis. While white blood cell counts at initial presentation can vary significantly, circulating lymphoblast cells are seen on peripheral blood smears in the majority of cases.

A bone marrow biopsy provides conclusive proof of ALL, typically with >20% of all cells being leukemic lymphoblasts. A lumbar puncture (also known as a spinal tap) can determine whether the spinal column and brain have been invaded. Brain and spinal column involvement can be diagnosed either through confirmation of leukemic cells in the lumbar puncture or through clinical signs of CNS leukemia as described above. Laboratory tests that might show abnormalities include blood count, kidney function, electrolyte, and liver enzyme tests.

Pathological examination, cytogenetics (in particular the presence of Philadelphia chromosome), and immunophenotyping establish whether the leukemic cells are myeloblastic (neutrophils, eosinophils, or basophils) or lymphoblastic (B lymphocytes or T lymphocytes). Cytogenetic testing on the marrow samples can help classify disease and predict how aggressive the disease course will be. Different mutations have been associated with shorter or longer survival. Immunohistochemical testing may reveal TdT or CALLA antigens on the surface of leukemic cells. TdT is a protein expressed early in the development of pre-T and pre-B cells, whereas CALLA is an antigen found in 80% of ALL cases and also in the "blast crisis" of CML.

Medical imaging (such as ultrasound or CT scanning) can find invasion of other organs commonly the lung, liver, spleen, lymph nodes, brain, kidneys, and reproductive organs.

A new method developed using data from the M.D. Anderson Cancer Center found that a haemoglobin level of 2.5 x 10/L, >0% immature myeloid cells, >10% bone marrow blasts causes a reduced overall survival. This data allows cases of CMML to be stratified into low, intermediate-1, intermediate-2 and high risk groups. These groups have median survival times of 24, 15, 8 and 5 months respectively.

Acute promyelocytic leukemia can be distinguished from other types of AML based on microscopic examination of the blood film or a bone marrow aspirate or biopsy as well as finding the characteristic rearrangement. Definitive diagnosis requires testing for the "PML/RARA" fusion gene. This may be done by polymerase chain reaction (PCR), fluorescent in situ hybridization (FISH), or conventional cytogenetics of peripheral blood or bone marrow. This mutation involves a translocation of the long arm of chromosomes 15 and 17. On rare occasions, a cryptic translocation may occur which cannot be detected by cytogenetic testing; on these occasions PCR testing is essential to confirm the diagnosis. Presence of multiple Auer rods on peripheral blood smear is highly suggestive of acute promyelocytic leukemia.

Definitive diagnosis of a chloroma usually requires a biopsy of the lesion in question. Historically, even with a tissue biopsy, pathologic misdiagnosis was an important problem, particularly in patients without a clear pre-existing diagnosis of acute myeloid leukemia to guide the pathologist. In one published series on chloroma, the authors stated that 47% of the patients were initially misdiagnosed, most often as having a malignant lymphoma.

However, with advances in diagnostic techniques, the diagnosis of chloromas can be made more reliable. Traweek et al. described the use of a commercially available panel of monoclonal antibodies, against myeloperoxidase, CD68, CD43, and CD20, to accurately diagnose chloroma via immunohistochemistry and differentiate it from lymphoma. Nowadays, immunohistochemical staining using monoclonal antibodies against CD33 and CD117 would be the mainstay of diagnosis. The increasingly refined use of flow cytometry has also facilitated more accurate diagnosis of these lesions.

Leukemia is diagnosed in a variety of ways. Some diagnostic procedures include:

- A bone-marrow aspiration and biopsy; marrow may be removed by aspiration or a needle biopsy.

- A complete blood count, which is a measurement of size, number, and maturity of different blood cells in blood.

- Blood tests may include blood chemistry, evaluation of liver and kidney functions, and genetic studies.

- A lymph-node biopsy; lymph node tissue is surgically removed to examine under a microscope, to look for cancerous cells.

- A spinal tap: a special needle is placed into the lower back into the spinal canal, which is the area around the spinal cord. Cerebral spinal fluid is fluid that bathes the child's brain and spinal cord. A small amount of cerebral spinal fluid is sent for testing to determine if leukemia cells are present.

Cytogenetic analysis has shown different proportions and frequencies of genetic abnormalities in cases of ALL from different age groups. This information is particularly valuable for classification and can in part explain different prognosis of these groups. In regards to genetic analysis, cases can be stratified according to ploidy, number of sets of chromosomes in the cell, and specific genetic abnormalities, such as translocations. Hyperdiploid cells are defined as cells with more than 50 chromosomes, while hypodiploid is defined as cells with less than 44 choromosomes. Hyperdiploid cases tend to carry good prognosis while hypodiploid cases do not. For example, the most common specific abnormality in childhood B-ALL is the t(12;21) "ETV6"-"RUNX1" translocation, in which the "RUNX1" gene, encoding a protein involved in transcriptional control of hemopoiesis, has been translocated and repressed by the "ETV6"-"RUNX1" fusion protein.

Below is a table with the frequencies of some cytogenetic translocations and molecular genetic abnormalities in ALL.

Following observation of the symptoms, the patients need to get complete blood counts and a bone marrow examination. If the patient has leukemia, the morphology and immunophenotype check is needed to make sure the type of leukemia.

The morphology of the blast in BAL is not certain. The cells could display both myeloid lineage and lymphoid or undifferentiated morphology. Therefore, the diagnosis cannot based on the morphology result. The immunophenotype check is the most important basis of the diagnosis of BAL.

Before 2008, the diagnosis of BAL was based on a score system proposed by the European Group for the Immunological Classification of Leukemias (EGIL) which could differentiate from other kinds of acute leukemia. The table shows this method.

If the score of only one lineage is higher than 2, the acute leukemia could be acute myeloid leukemia (AML) or acute lymphoblastic leukemia (ALL). According to the original EGIL scoring system BAL is defined when scores are over two points for both myeloid and T- or B- lymphoid lineages.

In 2008, WHO established a new and strict criteria standard for diagnosis of BAL. The presence of specific T-lymphoid antigens, cytoplasmic CD3 (cCD3), MPO and CD 19 became the most important standard for recognizing the lineage. Other B-lineage markers (CD22, CD79a, CD 10) and monocytic markers are also needed. Table 2 shows the method.

Compared with the EGIL scoring system, the current 2008 WHO criteria applied less but more specific markers to define the lineage of the blasts, and incorporated the intensity of markers expression into the diagnostic algorithm.

The diagnosis of BAL is so difficult that sometimes is misdiagnosed with AML or ALL because the morphology thus the therapy would not have a good effect.

Although not yet formally incorporated in the generally accepted classification systems, molecular profiling of myelodysplastic syndrome genomes has increased the understanding of prognostic molecular factors for this disease. For example, in low-risk MDS, "IDH1" and "IDH2" mutations are associated with significantly worsened survival.

Diagnosis is usually based on repeated complete blood counts and a bone marrow examination following observations of the symptoms. Sometimes, blood tests may not show that a person has leukemia, especially in the early stages of the disease or during remission. A lymph node biopsy can be performed to diagnose certain types of leukemia in certain situations.

Following diagnosis, blood chemistry tests can be used to determine the degree of liver and kidney damage or the effects of chemotherapy on the patient. When concerns arise about other damage due to leukemia, doctors may use an X-ray, MRI, or ultrasound. These can potentially show leukemia's effects on such body parts as bones (X-ray), the brain (MRI), or the kidneys, spleen, and liver (ultrasound). CT scans can be used to check lymph nodes in the chest, though this is uncommon.

Despite the use of these methods to diagnose whether or not a patient has leukemia, many people have not been diagnosed because many of the symptoms are vague, non-specific, and can refer to other diseases. For this reason, the American Cancer Society estimates that at least one-fifth of the people with leukemia have not yet been diagnosed.

The WHO 2008 classification of acute myeloid leukemia attempts to be more clinically useful and to produce more meaningful prognostic information than the FAB criteria. Each of the WHO categories contains numerous descriptive subcategories of interest to the hematopathologist and oncologist; however, most of the clinically significant information in the WHO schema is communicated via categorization into one of the subtypes listed below.

The WHO subtypes of AML are:

Acute leukemias of ambiguous lineage (also known as mixed phenotype or biphenotypic acute leukemia) occur when the leukemic cells can not be classified as either myeloid or lymphoid cells, or where both types of cells are present.

Flow cytometry is a diagnostic tool in order to count/visualize the amount of lymphatic cells in the body. T cells, B cells and NK cells are nearly impossible to distinguish under a microscope, therefore one must use a flow cytometer to distinguish them.

The prognosis for BAL patients is not good which is worse than ALL and AML. Medical Blood Institute reported cases of CR rate was 31.6%, with a median remission are less than 6 months

The median survival time is only 7.5 months. The life quality is also low because the immune function of patient is damaged seriously. They have to stay in hospital and need 24h care.

In another study, the results showed that young age, normal karyotype and ALL induction therapy will have a better prognosis than Ph+, adult patients. The study shows median survival of children is 139 months versus 11 months of adults, 139 months for normal karyotype patients versus 8 months for ph+ patients.

Blast crisis is the final phase in the evolution of CML, and behaves like an acute leukemia, with rapid progression and short survival. Blast crisis is diagnosed if any of the following are present in a patient with CML:

- >20% myeloblasts or lymphoblasts in the blood or bone marrow

- Large clusters of blasts in the bone marrow on biopsy

- Development of a chloroma (solid focus of leukemia outside the bone marrow)

There are two internationally accepted treatment protocols, which are geographically based:

- North America: the Children’s Oncology Group (COG) JMML study

- Europe: the European Working Group for Myelodysplastic Syndromes (EWOG-MDS) JMML study

The following procedures are used in one or both of the current clinical approaches listed above:

CML accounts for 8% of all leukaemias in the UK, and around 680 people were diagnosed with the disease in 2011.

The outlook in MDS is variable, with about 30% of patients progressing to refractory AML. The median survival rate varies from years to months, depending on type. Stem-cell transplantation offers possible cure, with survival rates of 50% at 3 years, although older patients do poorly.

Indicators of a good prognosis:

Younger age; normal or moderately reduced neutrophil or platelet counts; low blast counts in the bone marrow (< 20%) and no blasts in the blood; no Auer rods; ringed sideroblasts; normal or mixed karyotypes without complex chromosome abnormalities; and "in vitro" marrow culture with a nonleukemic growth pattern

Indicators of a poor prognosis:

Advanced age; severe neutropenia or thrombocytopenia; high blast count in the bone marrow (20-29%) or blasts in the blood;

Auer rods; absence of ringed sideroblasts; abnormal localization or immature granulocyte precursors in bone marrow section;

completely or mostly abnormal karyotypes, or complex marrow chromosome abnormalities and "in vitro" bone marrow culture with a leukemic growth pattern

Karyotype prognostic factors:

- Good: normal, -Y, del(5q), del(20q)

- Intermediate or variable: +8, other single or double anomalies

- Poor: complex (>3 chromosomal aberrations); chromosome 7 anomalies

The IPSS is the most commonly used tool in MDS to predict long-term outcome.

Cytogenetic abnormalities can be detected by conventional cytogenetics, a FISH panel for MDS, or virtual karyotype.

Prognosis refers to how well a patient is expected to respond to treatment based on their individual characteristics at time of diagnosis. In JMML, three characteristic areas have been identified as significant in the prognosis of patients:

Without treatment, the survival [5 years?] of children with JMML is approximately 5%. Only Hematopoietic Stem Cell Transplantation (HSCT), commonly referred to as a bone marrow or (umbilical) cord blood transplant, has been shown to be successful in curing a child of JMML. With HSCT, recent research studies have found the survival rate to be approximately 50%. Relapse is a significant risk after HSCT for children with JMML. It is the greatest cause of death in JMML children who have had stem cell transplants. Relapse rate has been recorded as high as 50%. Many children have been brought into remission after a second stem cell transplant.

Historically, hematological malignancies have been most commonly divided by whether the malignancy is mainly located in the blood (leukemia) or in lymph nodes (lymphomas).

However, the influential WHO Classification (published in 2001) placed a greater emphasis on cell lineage.

Relative proportions of hematological malignancies in the United States

Prognosis is generally good relative to other leukemias. Because of the acuteness of onset compared to other leukemias, early death is comparatively more common. The cause of early death is most commonly severe bleeding, often intracranial hemorrhage. Early death from hemorrhage occurs in 5-10% of patients in countries with adequate access to healthcare and 20-30% of patients in less developed countries. Risk factors for early death due to hemorrhage include delayed diagnosis, late treatment initiation, and high white blood cell count on admission. Despite advances in treatment, early death rates have remained relatively constant.

Relapse rates are extremely low. Most deaths following remission are from other causes, such as second malignancies, which in one study occurred in 8% of patients. In this study, second malignancies accounted for 41% of deaths, and heart disease, 29%. Survival rates were 88% at 6.3 years and 82% at 7.9 years.

In another study, 10-year survival rate was estimated to be approximately 77%.

Information on prognosis is limited by the rarity of the condition. Prognosis appears to be no different to AML in general, taking into account other risk factors. Acute erythroid leukemia (M6) has a relatively poor prognosis. A 2010 study of 124 patients found a median overall survival of 8 months. A 2009 study on 91 patients found a median overall survival for erythroleukemia patients of 36 weeks, with no statistically significant difference to other AML patients. AEL patients did have a significantly shorter disease free survival period, a median of 32 weeks, but this effect was explained by other prognostic factors. That is, AEL is often associated with other risk factors, like monosomal karyotypes and a history of myelodysplastic syndrome. Prognosis is worse in elderly patients, those with a history of myelodysplastic syndrome, and in patients who had previously received chemotherapy for the treatment of a different neoplasm.

Acute erythroid leukemia is rare, accounting for only 3–5% of all acute myeloid leukemia cases. One study estimated an occurrence rate of 0.077 cases per 100,000 people each year. 64–70% of people with this condition are male, and most are elderly, with a median age of 65.

For the analysis of a suspected "hematological malignancy", a complete blood count and blood film are essential, as malignant cells can show in characteristic ways on light microscopy. When there is lymphadenopathy, a biopsy from a lymph node is generally undertaken surgically. In general, a bone marrow biopsy is part of the "work up" for the analysis of these diseases. All specimens are examined microscopically to determine the nature of the malignancy. A number of these diseases can now be classified by cytogenetics (AML, CML) or immunophenotyping (lymphoma, myeloma, CLL) of the malignant cells.