SummaryIn transplant patients with impaired liver function, HPLC methodologies have been suggested for monitoring whole blood tacrolimus concentrations because of the reported inaccuracy of immunoassay for whole blood tacrolimus concentrations. One hundred fifty whole blood samples from 50 subjects enrolled in a multicenter liver transplant trial were chosen for HPLC/MS/MS analysis without consideration of the clinical status of the patient at the time of sampling. These samples were chosen to represent the sampling intervals during the 12-week posttransplantation period. Retrospectively, the authors identified a subset of 39 samples from 27 subjects exhibiting impaired liver function as demonstrated by bilirubin concentrations > 3.0 mg/dL (mean ± SD = 7.5 ± 5.6 mg/dL). The authors compared the agreement of concentrations obtained from the PRO-Trac II ELISA and HPLC/MS/MS by least squares linear regression analysis and Bland/Altman analysis, in this subset against the agreement of concentrations for 76 samples with normal bilirubin. In the samples obtained from patients with impaired liver function the resulting regression equation was: ELISA = 1.19(HPLC) + 0.7;r= 0.9. The mean difference (HPLC/MS/MS − ELISA) was −2.5 ng/mL ± 2.9 ng/mL (mean ± SD). While 71% of samples agreed within 3 ng/mL, 3% (n = 1) exhibited a difference >10 ng/ml. The corresponding evaluation of the samples with normal bilirubin concentrations resulted in the regression equation ELISA = 0.96(HPLC) + 0.9;r= 0.9, and a mean difference of −0.6 ng/mL ± 2.3 ng/mL. The authors conclude that while a small subset of patients with cholestasis may require closer evaluation with a more specific methodology, the majority of the patients may be satisfactorily monitored with the PRO-Trac II ELISA.