Background: Oxaliplatin-based chemotherapy can cause hepatic sinusoidal injury (HSI), portal hypertension, and splenic sequestration of platelets. Evidence suggests that bevacizumab may protect against HSI. Methods: Two cohorts of metastatic colorectal cancer (CRC) were analyzed: a nonrandomized exploratory cohort of 184 patients treated at a single institution from 2003 to 2010 and a confirmatory cohort of 200 patients from a multi-institutional randomized trial (NO16966). All patients were treated with frontline fluoropyrimidine and oxaliplatin with or without bevacizumab. Changes in splenic volumes and platelet counts were compared by treatment, two-sided log-rank test. Results: In the exploratory cohort, the bevacizumab-treated patients (n = 138) compared with the nonbevacizumab-treated patients (n = 46) demonstrated a longer median time to splenic enlargement (≥30%, P = .02) and reduced rate of thrombocytopenia (<150 000/mm3, P = .04). In the confirmatory cohort (106 bevacizumab arm and 94 placebo arm), the median time to a spleen enlargement of 30% or more was 7.6 vs 5.4 (P = .01), and six-month cumulative incidence of thrombocytopenia (platelets < 100 000/mm3) was 19% vs 51% (P < .001) for bevacizumab compared with placebo. The development of an increasing spleen size was associated with the risk of either grade 1 or grade 2 thrombocytopenia (P < .001). The cumulative rate of grade 1 or grade 2 thrombocytopenia was statistically less in the bevacizumab arm, with six-month grade 2 thrombocytopenia rates of 4% vs 23% (P < .001). Patients with a large spleen prior to chemotherapy initiation appeared to be at highest risk of this toxicity. Conclusion: In metastatic CRC, the addition of bevacizumab to oxaliplatin-based chemotherapy reduces the frequency of splenic enlargement and the rate of thrombocytopenia.