Temozolomide has shown efficacy in the treatment of metastatic melanoma similar to that of dacarbazine (DTIC), the standard chemotherapy, but with the added benefit of penetration into the central nervous system (CNS). Isolated CNS relapse is increasingly a problem for patients who respond to biochemotherapy. By replacing DTIC with temozolomide in treatment regimens, the incidence of CNS relapse might be reduced. This hypothesis is difficult to test in a prospective randomized controlled trial because of the large number of patients that would be required. We have examined this question in a retrospective case control study, observing the rates of CNS relapse in advanced metastatic melanoma patients responding to DTIC- or temozolomide-based chemotherapy in three institutions. Twenty-one DTIC and 20 temozolomide responders were identified, and have been followed up for a median of 19.0 months (range 6.0-74.3 months). CNS relapse occurred in nine DTIC- and two temozolomide-treated patients, a statistically significant difference in favour of the new agent (P = 0.03). These results support the investigation of temozolomide as a replacement for DTIC in systemic treatment regimens for melanoma.