Summary. A randomized trial has been performed in which 91 patients with stage III myeloma and additional severe criteria were randomly allocated to either VAD or VMBCP. No significant difference was noted between these two groups using the following criteria: response rate (VMBCP: 54%; VAD: 39%), impact on symptoms, median survival (VMBCP: 14 months, VAD: 17 months). However, toxic effects and refusal to pursue treatment were more frequent with VAD than with VMBCP (12 v 6). Therefore, in this trial, VMBCP appears to be more useful than VAD.