The response to treatment with low-dose ara-C was studied in 102 consecutive patients; 79 with myelodysplastic syndrome (MDS) and 23 with acute myelogenous leukaemia (AML) following MDS. The aim was to find variables that could predict the response to treatment. All patients had clinical symptoms related to cytopenia. Peripheral blood values, bone marrow morphology histology and chromosomes were analysed before the start of treatment. The median survival of the patients was 9 months and a poor survival was predicted by advanced age, low platelet counts, the presence of pseudo-Pelger morphology and > or = 2 chromosomal aberrations. Thirty patients (29%) responded with either a complete remission or a significant increase in haemoglobin level. For the remaining 71%, the treatment was ineffective and in some cases hazardous. The factors associated with a poor response to treatment could be divided into two groups: one included low platelet counts and the presence of chromosomal aberrations, both signs of progressive MDS with a short survival, and the other comprised morphological findings, indicating ineffective haemopoiesis. Patients with platelet counts > 150 x 10(9)/l had a response rate of 55% compared to 23.5% in patients with subnormal platelet counts. Logistic regression identified low bone marrow cellularity, absence of ring sideroblasts and < 2 chromosomal aberrations as predictors of a favourable response in patients with platelet counts < 150 x 10(9)/l. These factors and the platelet count were combined in a predictive model which can divide patients into three groups with different probabilities of response: a favourable group, 38.6% of the patients, with a response rate of > 50%, an intermediate group, 32.7% of the patients, with a response rate of 24%, and an unfavourable group, 28.7% of the patients, with only 3% responses. While low-dose ara-C is an effective treatment for some patients, it is ineffective and hazardous for others. We present a model that can facilitate therapeutic decision making in two-thirds of patients with MDS and MDS-AML by identifying patients who should not be treated with low-dose ara-C as well as patients with a relatively high probability of response.