Despite a widespread belief in both the academic and public policy literatures that working long hours is deleterious to health and well-being, our critical review of this large and complex literature fails to support a robust direct causal effect of work hours on either physical or mental well-being outcomes. Large-scale epidemiological studies, many of which are prospective and include objective health outcome measures, support a statistically significant association between long work hours and coronary heart disease and depression, but the effect sizes are very small. Moreover, there is an absence of true longitudinal studies that assess the consistency of working long hours over time and its relationship to well-being. Our review suggests that the effects of working long hours are nuanced in that they may vary considerably for different working populations based on gender, age, working conditions, and other factors. Primary and meta-analytic studies suggest that such moderator effects are plausible, yet rigorous testing of these remains to be done. We conclude with suggestions for specific moderator effects that seem worth investigating in future research.