Objective: Headache in employees may be linked with both overwork and sleep restriction induced by long working hours. Inter-relationships among working hours, sleep duration and headache were investigated. Methods: Cross-sectional analyses for prevalent headache (n = 35,908) and 1-year follow-up analyses for incident headache (n = 19,788) were conducted in apparently healthy white-collar men aged 25–59 years. Headache (yes/no), working hours and sleep duration were based on self-administered questionnaire. After determination of relationships between working hours and sleep duration, logistic regression analysis estimated odds ratio (OR) and 95% confidence interval for prevalent and incident headache according to working hours (35–44, 45–49, 50–59 and ≥60 h/week) and sleep duration (≥7, 6–6.9, 5–5.9 and &lt
5 h/day), and tested linear trends in OR. Additionally, interactive effects of working hours and sleep duration on OR were checked. Covariates in the analyses were age, body mass index, drinking, smoking and exercise. Results: Prevalent and incident headache was found in 1979 (5.5%) men and 707 (3.6%) men, respectively. Working hours were inversely associated with sleep duration. OR for prevalent and incident headache rose with increasing working hours and with reducing sleep duration, regardless of influences of the covariates. Working hours and sleep duration had no interactive effects on OR for prevalent or incident headache. Conclusions: The results indicate that long working hours directly and indirectly (via short sleep duration) induce headache even in apparently healthy white-collar men. Headache in employees may be useful for early detection of adverse health effects by long working hours.