This study aimed to verify whether the advantage in terms of response rate and survival of dacarbazine plus tamoxifen over dacarbazine alone in metastatic malignant melanoma reported in a previous randomized trial was due to a specific interaction of dacarbazine with tamoxifen. A total of 125 patients with locoregional or disseminated malignant melanoma were randomized to receive dacarbazine (250 mg/m(2) days 1-5 every 3 weeks) plus tamoxifen (arm A) or vindesine (3 mg/m(2) every week for 6 weeks, then every 2 weeks) plus tamoxifen (arm B). Of the 125 randomized patients, 57 and 59 were evaluable in arm A and B, respectively. The complete response rates were the same (2% versus 2%) and the complete plus partial response rates were similar (11% versus 14%) in the two groups. There was no significant difference in survival. Neither response or survival correlated with gender. In conclusion, when combined with tamoxifen, dacarbazine does not have a specific effect on response or survival compared with vindesine. The lower response rate to dacarbazine plus tamoxifen (11%) than that reported in the previous trial (28%) might be explained by actual differences in patient and/or participating centre accrual characteristics in the presence of apparently identical eligibility criteria.