Myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) are heterogeneous neoplasms of hematopoietic stem cells that are challenging to diagnose, differentiate, and prognosticate. Cytogenetic and mutational analyses are useful in humans but unavailable for dogs, where diagnosis and classification still rely largely on hematologic and morphologic assessment. The objectives of this study were to apply a classification scheme to myeloid neoplasms and to assess outcome in relation to predictor variables. Keyword search of a laboratory database, application of sequential exclusion criteria, and consensus from 3 reviewers yielded 70 cases of myeloid neoplasia with hematology results, and cytologic (11), histologic (14), or both (45) types of marrow specimens. Based on blast percentage and morphology, 42 cases were classified as MDS and 28 as AML. Dogs with MDS had significantly lower body weights, hemoglobin concentrations and blood blasts, and higher red blood cell size variability and platelet numbers than dogs with AML. Estimates of median survival using Kaplan-Meier curves for dogs with MDS and AML were 384 and 6 days, respectively (P < .001). The instantaneous risk of death for dogs with MDS was approximately 5× lower than that of dogs with AML. Significant predictor variables of survival were body weight, white blood cell count, platelet count, and percent blood blasts (P < .05). Hazard ratios (HRs) derived from best-fitting Cox regression models were 1.043, 0.998, and 1.061 for increased neutrophils, decreased platelets, and increased blood blasts, respectively. Findings from this study suggest that hematologic and morphologic variables are useful to predict outcomes in myeloid neoplasia.