Several cases of cardiac adverse reactions related to vinorelbine (VNR) have been reported in the literature. In order to quantify the incidence of these cardiac events, we performed a meta-analysis of clinical trials comparing VNR with other chemotherapeutic agents in the treatment of various malignancies. Randomized clinical trials comparing VNR with other drugs in the treatment of cancer were searched in Medline, Embase, Evidence-based Medicine Reviews databases and the Cochrane library from 1987 to 2002. Outcomes of interest were severe cardiac events, toxic deaths and cardiac event-related deaths reported in each publication. We found 19 trials, involving 2441 patients treated by VNR and 2050 control patients. The incidence of cardiac events with VNR was 1.19% [95% confidence interval (CI) (0.75; 1.67)]. There was no difference in the risk of cardiac events between VNR and other drugs [odds ratio: 0.92, 95% CI (0.54; 1.55)]. The risk of VNR cardiac events was similar to vindesine (VDS) and other cardiotoxic drugs [fluorouracil, anthracyclines, gemcitabine (GEM) em leader ]. Even if it did not reach statistical significance because of a few number of cases, the risk was lower in trials excluding patients with cardiac history, and seemed to be higher in trials including patients with pre-existing cardiac diseases. Vinorelbine-related cardiac events concern about 1% of treated patients in clinical trials. However, the risk associated with VNR seems to be similar to that of other chemotherapeutic agents in the same indications.