Background <p>Some epidemiological studies suggest that exposure to bisphenol A is associated with diabetes; however, there is limited evidence of a dose-response relationship between urinary benzophenone-3 (BP-3) and diabetes. The aim of this study was to investigate the association between urinary BP-3 levels and the risk of diabetes in American adults.</p> Methods <p>A cross-sectional study utilized data from the National Health and Nutrition Examination Survey (NHANES) to assess participants aged 20 and older in the United States from 2005 to 2016. The exposure variable was urinary BP-3 and the outcome diabetes risk. Multivariate logistic regression and subgroup analyses were performed to assess the independent association between urinary BP-3 and diabetes risk. Smooth curve fits were applied to examine the non-linear relationship. We performed several sensitivity analyses to evaluate the robustness of our results.</p> Results <p>Our study included 8027 participants, 50.3% of whom were male, and with a mean age of 49.3 ± 17.3 years. After adjustment for potential confounders, the analysis indicated that each unit increase in log₂-transformed BP-3 was associated with a 5% reduction in diabetes risk (OR = 0.95, 95% CI: 0.93–0.97, <i>P</i> &lt; 0.001). Compared to the Q1 group, the risk of developing diabetes was significantly lower in the Q4 group (OR = 0.63, 95% CI: 0.52–0.78, <i>P</i> &lt; 0.001). L-shaped relationship was observed between log₂ BP-3 and diabetes in adults in the USA, with an inflection point of 6.801 ng/ml (OR = 0.916, 95% CI: 0.885–0.949, <i>P</i> &lt; 0.001). Our sensitivity analyses confirmed the stability of our results.</p> Conclusions <p>Urinary BP-3 levels exhibited an L-shaped relationship with diabetes risk in the United States, with an inflection point at approximately 6.801 ng/mL.</p> Clinical trial number <p>Not applicable.</p>