To examine the clinical significance of clotrimazole troche discontinuation on tacrolimus trough levels and risk of allograft rejection after pancreas transplantation. Retrospective cohort study. Academic medical center. Sixty-five pancreas transplant recipients (simultaneous pancreas-kidney transplants [39 patients], pancreas after kidney transplants [4 patients], and pancreas transplant alone [22 patients]) who were discharged after transplantation receiving a maintenance immunosuppressive regimen of tacrolimus, mycophenolate, and prednisone, and a clotrimazole troche to prevent oral mucosal candidiasis; per protocol, the clotrimazole troche was discontinued at 3 months after transplantation. Patients were followed for 1 year after transplantation. The primary outcome measure was the difference in tacrolimus trough level before and after discontinuation of the clotrimazole troche. The secondary outcome measure was the difference in tacrolimus trough level when patients were stratified by the cohort that had a documented rejection episode 3-12 months after transplantation (rejection group) and the cohort that did not experience a rejection episode (no-rejection group). The incidence of rejection was evaluated in relation to mean tacrolimus trough concentrations above or below a protocol-defined level of significance (6 ng/ml). For the primary outcome, the mean tacrolimus trough level before discontinuation of the clotrimazole troche was significantly higher than the mean trough level after discontinuation (mean ± SD 9.6 ± 3.0 ng/ml vs 7.1 ± 2.6 ng/ml, p = 0.000003). For the secondary outcome, the mean tacrolimus trough level difference before and after clotrimazole troche discontinuation remained significant in both the no-rejection group (9.5 ± 3.0 ng/ml vs 7.4 ± 2.4 ng/ml, p = 0.00007) and rejection group (10.9 ± 3.3 ng/ml vs 4.1 ± 2.5 ng/ml, p = 0.0008). Between groups, the mean tacrolimus serum trough level after clotrimazole troche discontinuation was lower in the rejection group (4.1 ± 2.5 ng/ml) than that in the no-rejection group (7.4 ± 2.4 ng/ml; p = 0.005). The mean tacrolimus trough level difference between before and after discontinuation was greater in the rejection group (6.8 ± 1.5 ng/ml) versus the no-rejection group (2.1 ± 3.8 ng/ml, p = 0.009). Tacrolimus trough levels below 6 ng/ml (19 patients) after clotrimazole troche discontinuation were associated with an increased incidence of rejection episodes within 3-12 months after transplantation (odds ratio 12, 95% confidence interval 1.24-115.91, p = 0.032) versus trough levels of 6 ng/ml or higher (46 patients). Clotrimazole troche discontinuation at 3 months after transplantation may cause significant tacrolimus trough level reductions. In addition, when trough levels are below 6 ng/ml, these fluctuations may contribute to the occurrence of allograft rejection.