Purpose To determine whether adding bevacizumab, with or without mitomycin, to capecitabine monotherapy improves progression-free survival (PFS) in patients with metastatic colorectal cancer (mCRC) in an open-label, three-arm randomized trial. Patients and Methods Overall, 471 patients in Australia, New Zealand, and the United Kingdom with previously untreated, unresectable mCRC were randomly assigned to the following: capecitabine; capecitabine plus bevacizumab (CB); or capecitabine, bevacizumab, and mitomycin (CBM). We compared CB with capecitabine and CBM with capecitabine for progression-free survival (PFS). Secondary end points included overall survival (OS), toxicity, response rate (RR), and quality of life (QOL). Results Median PFS was 5.7 months for capecitabine, 8.5 months for CB, and 8.4 months for CBM (capecitabine v CB: hazard ratio [HR], 0.63; 95% CI, 0.50 to 0.79; P < .001; C v CBM: HR, 0.59; 95% CI, 0.47 to 0.75; P < .001). After a median follow-up of 31 months, median OS was 18.9 months for capecitabine and was 16.4 months for CBM; these data were not significantly different. Toxicity rates were acceptable, and all treatment regimens well tolerated. Bevacizumab toxicities were similar to those in previous studies. Measures of overall QOL were similar in all groups. Conclusion Adding bevacizumab to capecitabine, with or without mitomycin, significantly improves PFS without major additional toxicity or impairment of QOL.