Background and objectives:Nonadherence to oral immunosuppressive drugs in renal transplant patients remains a major challenge. The objective of this study was to develop an adherence-exposure model that 1) quantifies the impact of nonadherence patterns on cyclosporine levels and 2) identifies nonadherence patterns that are associated with unfavorable transplantation outcomes.Design, setting, participants, and measurements:This model quantified variability in drug exposure, expressed as the coefficient of variation (CV%), for time-averaged and trough cyclosporine levels (Cavg and Cmin, respectively), and percentage of days spent below the therapeutic Cmin target. Simulated patterns of nonadherence closely matched those observed in clinical practice for four nonadherence clusters and an “Others” category.Results:Patients in simulated nonadherence clusters 1–3 spent a mean (standard deviation) 5.8% (4.9), 9.0% (5.0), and 6.5% (3.4) of days below the Cmin target, compared with 76.8% (6.5) for cluster 4 and 38.3% (6.4) for the “Others” category. Mean (standard deviation) CV% values for Cmin were 24.1 (7.9), 35.4 (11.7), and 34.1 (10.6) for clusters 1–3, compared with 136.4 (23.6) for cluster 4 and 64.8 (10.3) for the “Others” category. Findings for Cavg were similar.Conclusion:Based on nonadherence patterns and known relationships between CV% for Cmin and Cavg, and transplantation outcomes, patients in cluster 4 and the “Others” category are expected to be at high risk of allograft rejection. The proposed drug adherence-exposure model is useful to identify high-risk patients who can be targeted for interventions aimed at enhancing drug adherence to optimize clinical long-term outcomes.