Background: Optimal platelet inhibition is essential for minimizing both thrombotic and hemorrhagic complications in patients with coronary artery disease (CAD). Although high on-treatment platelet reactivity (HPR) has been consistently associated with adverse clinical outcomes, the relationship between platelet reactivity—measured as P2Y12 reaction units (PRU)—and cardiovascular mortality remains incompletely characterized. In particular, potential non-linear associations have not been adequately explored. Objective: We aimed to investigate the association between PRU and cardiovascular mortality in patients with CAD, with a specific focus on identifying potential non-linear relationships. Methods: We conducted a retrospective observational cohort study including 1000 patients with angiographically confirmed CAD treated at a tertiary cardiology center in Almaty, Kazakhstan, between 2024 and 2025. Platelet reactivity was assessed using the VerifyNow P2Y12 assay. Multivariable logistic regression models were used to identify independent predictors of cardiovascular mortality. To assess potential non-linear associations between PRU and mortality, restricted cubic spline regression was applied with predefined knot placement. Model performance was evaluated in terms of discrimination (C-statistic) and calibration (Hosmer-Lemeshow goodness-of-fit test). Results: In conventional linear regression models, PRU was not independently associated with cardiovascular mortality (odds ratio [OR] ~1.00; p > 0.05). However, spline-based analyses demonstrated a statistically significant non-linear (U-shaped) relationship between PRU and mortality risk (p for non-linearity = X). Both low and high PRU values were associated with increased mortality, whereas intermediate PRU levels corresponded to the lowest observed risk. Additional independent predictors of mortality included advanced age, diabetes mellitus, and elevated inflammatory markers. Conclusions: Our findings reveal a significant non-linear association between platelet reactivity and cardiovascular mortality in patients with CAD. Both insufficient and excessive platelet inhibition appear to confer increased risk, suggesting that optimal PRU targets may lie within an intermediate therapeutic range. These results support a paradigm shift toward more individualized antiplatelet therapy strategies guided by platelet function testing.