To investigate whether a relationship between chemotherapy-associated adverse events and treatment efficacy exists, we have analysed the toxicity, objective response and survival data of 303 patients with advanced colorectal cancer. Patients were divided into two groups: the first with beneficial effect (I, n = 245), and the second with progressive disease (II, n = 58). Differences in terms of incidence rates, type and severity of ad verse events were analysed with univariate and multivariate models. The median number of side effects in group I was 6 vs 4 in group II (OR=1.342; P= 0.0001). An inverse correlation between disease control and treatment tolerance was confirmed when side effects were analysed according to severity and type of treatment-associated toxicities (haematological: P = 0.0005 vs nonhaematological P = 0.0001). When median survival was analysed according to the number of adverse events, it was 10 (95% CI, 3-7), 16 (14-18), and 18 (16-20) months in case of 0-1, 2-5, and > or =6 adverse events, respectively (P = 0.01). In conclusion, the results of this analysis suggest that occurrence of side effects during chemotherapy in advanced colorectal cancer is an independent and reliable prognostic indicator for response and survival.