Objectives: To examine the impact of sleep health on work in Japan and to investigate the relationship between presenteeism and sleep health. Design: Cross sectional analysis of a questionnaire survey. Setting: Seventeen offices in Tokyo, Japan. Participants: The study included 2897 participants, of which 1835 were men and 1062 women, aged between 18 and 76 years. Measurements: Productivity loss was measured using the Short Form of the Work Limitations Questionnaire (WLQ-SF). Results: The decline in productivity due to presenteeism of people with short sleep duration, between 5 and 6 h (beta = 0.068, p = 0.004) and less than 5 h (beta = 0.105, p < 0.001), was significantly greater compared with those with long sleep duration (7-8 h). Subjective sleep quality (beta = 0.124, p < 0.001), sleep latency (beta = 0.073, p < 0.001), sleep disturbance (beta = 0.123, p < 0.001), use of sleep medication (beta = 0.044, p = 0.007), and daytime dysfunction (beta = 0.359, p = 0.001) significantly affected presenteeism. When adjusting for confounding factors, sleep duration on workdays, sleep duration on free days, mid-sleep on free days corrected for sleep debt on workdays (MSFsc), and social jet lag were not associated with presenteeism in the multiple regression analysis. Conclusions: Sleep health is associated with presenteeism in Japan. Subjective sleep quality, daytime dysfunction, sleep disturbance, and use of sleep medicine are associated with presenteeism. Good sleep hygiene may be important for workers' productivity. (C) 2020 National Sleep Foundation. Published by Elsevier Inc. All rights reserved.