The disruption of the circadian timing system (CTS), which rhythmically controls cellular metabolism and proliferation, accelerated experimental cancer progression. A measure of CTS function in cancer patients could thus provide novel prediction information for outcomes, and help to identify novel specific therapies. The rest-activity circadian rhythm is a reliable and non-invasive CTS biomarker, which was monitored using a wrist watch accelerometer for 2 days in 436 patients with metastatic colorectal cancer. The relative percentage of activity in-bed versus out-of-bed (I < O) constituted the tested CTS measure, whose prognostic value for overall survival (OS) and progression-free survival (PFS) was determined in a pooled analysis of three patient cohorts with different treatment exposures. Median OS was 21.6 months [17.8–25.5] for patients with I < O above the median value of 97.5% as compared to 11.9 months [10.4–13.3] for those with a lower I < O (Log-rank p < 0.001). Multivariate analyses retained continuous I < O as a joint predictor of both OS and PFS, with respective hazard ratios (HR) of 0.954 (p < 0.001) and 0.970 (p < 0.001) for each 1% increase in I < O. HRs had similar values in all the patient subgroups tested. The circadian physiology biomarker I < O constitutes a robust and independent quantitative predictor of cancer patient outcomes, that can be easily and cost-effectively measured during daily living. Interventional studies involving 24-h schedules of clock-targeted drugs, light intensity, exercise and/or meals are needed for testing the relevance of circadian synchronization for the survival of patients with disrupted rhythms.