Fludarabine is effective in chronic lymphocytic leukaemia (CLL) and low-grade non-Hodgkin's lymphoma (NHL). A major side-effect of this purine analogue is immunosuppression which may favour opportunistic infections. Additionally, impairment of immunosurveillance might promote Epstein-Barr virus (EBV) reactivation and possibly favour transformation to high-grade malignancy. The aim of this study was to evaluate the immunosuppression-related effects of the fludarabine-based combination Flucyd in advanced low-grade NHL or CLL by serially monitoring T-lymphocyte subsets, opportunistic infections, EBV-reactivation, and histologic transformation. 24 patients with advanced NHL (n = 21) or CLL (n = 3) received fludarabine 25 mg/m2/d + cyclophosphamide 350 mg/m2/d + dexamethasone 20 mg/d in 3 d courses for a maximum of six courses. The overall response rate was 79% (eight CR, 11 PR, five failures); 11 patients relapsed or progressed between 3 and 19 months from response, and eight are in CR or PR at 3-27 months. The CD4+ lymphocyte counts decreased significantly during therapy from a median of 484/microliter pre-treatment (range 142-1865) to a median of 198/microliter (71-367). In 19 responders monitored off therapy every 3 months until relapse/progression, CD4+ counts were persistently low with minimal recovery over time. During treatment, 16 infections occurred in 11/24 patients. No delayed opportunistic infections occurred in responders while off therapy. The circulating EBV DNA load serially measured in 19 patients by a quantitative PCR assay showed an increase in four patients during treatment. A lymph node biopsy performed in two of these was PCR positive for EBV DNA, whereas LMP1 and EBERs were negative. Six NHL patients evolved into high-grade B-cell NHL. In conclusion, fludarabine combined with cyclophosphamide and dexamethasone is an effective therapy for recurrent indolent lymphoma. This combination produces prolonged T-lymphocytopenia and has the potential to reactivate a latent EBV infection. T-cell dysfunction, however, is not associated with higher incidence of clinical opportunistic infections and does not adversely influence clinical outcome.