This prospective cohort study was conducted to assess the duration of daytime napping and its effect combined with night sleep deprivation on the risk of developing high HOMA-IR (homeostasis model assessment of insulin resistance) index and disadvantageous changes in glycosylated hemoglobin (HbA1c) levels.A total of 5845 diabetes-free subjects (2736 women and 3109 men), 30 to 65 years of age, were targeted for this cohort study since 2008. Multiple adjusted Cox regression models were performed to evaluate the single and joint effects of daytime napping on the risk of an elevated HbA1c level and high HOMA-IR index.After an average of 4.5 years of follow-up, >30 minutes of daytime napping was significantly associated with an increased risk of an elevated HbA1c level (>6.5%) in men and women (all P trend < 0.05). Hazard ratios (HRs) for an HbA1c level between 5.7% and 6.4% were also significant in the entire cohort and women, but nonsignificant in men. HRs (95% confidence interval, CIs) for the high HOMA-IR index in the entire cohort, men, and women were 1.33 (1.10-1.62), 1.46 (1.08-1.98), and 1.47 (1.12-1.91), respectively. The combination of sleep deprivation with no naps or >30 minutes napping and the combination of no sleep deprivation with >30 minutes daytime napping were all associated with an HbA1c level >6.5% (HR = 2.08, 95% CI = 1.24-3.51; HR = 4.00, 95% CI = 2.03-7.90; and HR = 2.05, 95% CI = 1.29-3.27, respectively). No sleep deprivation combined with >30 minutes daytime napping correlated with a high risk of an HbA1c level between 5.7% and 6.4% and high HOMA-IR index (HR = 2.12, 95% CI = 1.48-3.02; and HR = 1.35, 95% CI = 1.10-1.65, respectively).Daytime napping >30 minutes was associated with a high risk of an elevated HbA1c level and high HOMA-IR index. No sleep deprivation combined with napping >30 minutes carries a risk of abnormal glucose metabolism. Sleep deprivation combined with brief daytime napping <30 minutes was not associated with a risk for an elevated HbA1c level and high HOMA-IR index.