<p>Global warming has increased the risk of vector-borne diseases, leading to more widespread pesticide use. While health effects of occupational or agricultural pesticide exposure are well documented, evidence linking residential pesticide use to mortality remains limited. We examined the association between residential pesticide exposure in U.S. adults, urinary pesticide metabolites, and both all-cause and cancer-specific mortality to assess the risks of household pesticide use. Data were obtained from the National Health and Nutrition Examination Survey (NHANES). Weighted multivariable Cox regression estimated hazard ratios (HRs) for residential pesticide exposure and mortality. Mediation analysis evaluated the role of urinary metabolites. The dataset was divided into training and validation sets, and three predictive models were developed using machine learning. The study included 21,466 adults. Compared to unexposed participants, those reporting residential pesticide exposure had HRs of 1.444 (95% CI, 1.239–1.683) for all-cause mortality and 1.375 (95% CI, 1.011–1.870) for cancer-specific mortality. Urinary levels of N,N-diethyl-meta-toluamide and N,N-diethyl-3-hydroxymethyl benzamide statistically mediated 20.51% and 19.01% of these risks. All models showed strong predictive performance, with AUC values &gt; 0.8 in both training and validation sets. Residential pesticide exposure may be associated with an increased mortality risk, partially mediated by urinary metabolite levels. Limiting unnecessary household pesticide exposure may help reduce potential health risks, while the predictive model developed in this study may support future risk stratification and the identification of individuals at higher risk, pending external validation.</p>