The objective of the study was to determine the effectiveness and the toxicity of a combined chemotherapy consisting of cladribine (2-CdA), mitoxantrone and cyclophosphamide (CMC regimen) in the treatment of previously untreated B cell chronic lymphocytic leukemia (B-CLL). From August 1998 to December 2000 2-CdA was administered at a dosage of 0.12 mg/kg for 3 (CMC3) or 5 (CMC5) consecutive days, mitoxantrone at 10 mg/m2 on day 1 and cyclophosphamide at 650 mg/m2 on day 1 to 62 patients with advanced or progressive B-CLL. The cycles were repeated at 4 week intervals or longer if severe myelosuppression occurred. Twenty patients received CMC5 and 42 patients CMC3. Within the analyzed group an overall response (OR) rate (CR+PR) of 64.5% (95% CI: 52.7-76.3%) was reported, including 29.0% CR. There was no difference in the CR rate between the patients treated with CMC5 (30%) and CMC3 (28.6%) (P = 0.9), nor in the OR rate (55.0% and 69.0%, respectively, P = 0.3). Residual disease was identified in seven out of 18 (38.9%) patients who were in CR, including two treated with CMC5 and five treated with CMC3 protocols. CMC-induced grade III or IV thrombocytopenia occurred in 12 (19.4%) of patients, including four (20%) CMC5-treated and eight (19%) CMC3-treated patients (P= 0.8). Neutropenia grade III or IV was observed in seven (35%) and 11 (26.2%) patients, respectively (P = 0.8). Severe infections, including pneumonia and sepsis, occurred more frequently after CMC5 (11 patients, 55.0%) than CMC3 (10 patients, 28.6%) (P = 0.03) Fourteen patients died, including six treated with CMC5 and eight treated with CMC3 (30% and 19%, respectively). Infections were the cause of death in nine patients, including four in the CMC5 group and five in the CMC3 group. In conclusion, our results indicate that the CMC programme is an active combined regimen in previously untreated B-CLL patients; its efficiency seems to be similar to that observed earlier in B-CLL patients treated with 2-CdA as a single agent. However, toxicity, especially after CMC5 administration, is significant. Therefore, we recommend the CMC3 but not the CMC5 programme for further evaluation.