To examine the association between urinary metabolites of polycyclic aromatic hydrocarbons (OH-PAHs) and diabetes, online databases, including PubMed, Scopus, and Web of Science, were searched on July 17, 2019. Of the 668 articles identified through searching, six cross-sectional studies involving 24,406 participants were included. The pooled odds ratio (OR) and 95% confidence interval (CI) were calculated using a random-effect model. Heterogeneity was measured by reporting the I-square index. Moreover, subgroup analysis according to types of metabolites was performed. We found a significantly higher odds of diabetes in the highest versus the lowest category of urinary naphthalene (NAP), fluorine (FLU), phenanthrene (PHEN), and total OH-PAH metabolites. The pooled OR (95% CI) was estimated at 1.47 (1.17, 1.78), 1.50 (1.29, 1.71), 1.41 (1.21, 1.60), and 1.61 (1.01, 2.21), respectively. We also found a significant association per 1-fold increase in FLU (OR = 1.09, 95% CI [1.00, 1.19]) and PHEN (OR = 1.19, 95% CI [1.08, 1.30]) metabolites. In subgroup analysis stratified by types of OH-PAH metabolites, A significant stronger odds of diabetes was observed in the highest versus the lowest category of 2-PHEN (OR = 1.66, 95% CI [1.32, 2.00]), 2-NAP (OR = 1.66, 95% CI [1.16, 2.17]), 2-FLU (OR = 1.62, 95% CI [1.28, 1.97]), and 9-FLU (OR = 1.62, 95% CI [1.21, 2.04]) metabolites. Furthermore, there was a meaningfully greater likelihood of diabetes per 1-fold increase in 2-FLU (OR = 1.34, 95% CI [1.10, 1.57]), 2-PHEN (OR = 1.33, 95% CI [1.14, 1.51]), and 3-PHEN (OR = 1.19, 95% CI [1.04, 1.34]) metabolites. In conclusion, our study suggests the significant odds of association between urinary OH-PAH metabolites and diabetes. (C) 2020 Elsevier Ltd. All rights reserved.