Renewed interest in chronic lymphocytic leukemia (CLL) has led to an unprecedented number of investigators contributing to all aspects of research in this disease. In fact, the evolution of research in the area of molecular aberrations in CLL and their impact on treatment resistance alone is striking. These data, along with the advent of the purine analogs, have been central to this paradigm shift. The inferior response rate, the abbreviated response duration, and the inability to prolong survival with alkylating agents such as chlorambucil have resulted in purine analogs being used as first- and second-line therapy for patients with CLL. In fact, patients treated with fludarabine have a higher overall and complete response rate as well as a disease-free survival advantage compared with patients treated with alkylator-based therapy. Pentostatin, the first purine analog to enter clinical trials, was never subjected to extensive schedule optimization despite its demonstrated efficacy and its paucity of significant myelosuppression compared with the other purine analogs. However, pentostatin induced a 25% to 30% response rate in heavily pretreated CLL patients, including some who had received prior fludarabine, suggesting possible non-cross-resistance. Based on preclinical data demonstrating synergistic activity when a DNA damaging agent (eg, an alkylating agent) is followed by an inhibitor of DNA repair (a purine analog), a number of purine analog/alkylator combinations have been and are presently being examined in a variety of lymphoid neoplasms. While the clinical data conflict, at least two phase II studies examining a combination of a purine analog and an alkylator in untreated patients with CLL have generated promising data. This report describes the scientific justification and the design of a new phase II study examining the combination of pentostatin and chlorambucil with granulocyte-macrophage colony-stimulating factor support for patients with untreated, treated, and fludarabine-refractory B-cell CLL.