This report constitutes a prospectively planned meta-analysis combining two almost identical trials undertaken in Australasia and Canada to study the effect of starting chemotherapy immediately in asymptomatic patients with metastatic colorectal cancer. Patients (n=168) were randomised to receive either immediate or delayed treatment (at onset of predefined symptoms). Australasian patients received either weekly 5-fluorouracil and leucovorin (500 and 20 mg m(-2), respectively) (n=59) or the daily x 5 Mayo Clinic schedule (425 and 20 mg m(-2), respectively) (n=42). Canadian patients were treated with the Mayo schedule (n=67). Otherwise, the two studies were almost identical in design and each used the European Organisation for the Research and Treatment of Cancer (EORTC) QLQ-C30 instrument for measuring quality of life (QoL). Treatment was continued until 6 months had elapsed or disease progression occurred. Low accrual led to trial suspension before the predetermined sample size for either study was reached. Median survival was not significantly better with immediate treatment (median 13.0 vs 11.0 months; hazard ratio, 1.15; 95% confidence interval (CI) 0.79-1.72; P=0.49). There was no statistically significant difference in progression-free survival (time from randomisation until first evidence of progression after chemotherapy, 10.2 vs 10.8 months; hazard ratio, 1.08; 95% CI 0.71-1.64; P=0.73). There was no difference in overall QoL or its individual domains between the two treatment strategies at baseline or at any subsequent time point. Early treatment of asymptomatic patients with metastatic colorectal cancer did not provide a survival benefit or improved QoL compared to withholding treatment until symptoms occurred.