FK506 (tacrolimus), a macrolide immunosuppressant, is widely used in pediatric transplant patients, but a relatively narrow therapeutic window in children vs adults requires close and accurate monitoring of whole blood FK506 levels. High-pressure liquid chromatography/tandem mass spectrometry (HPLC/MS/MS)-based assays have been viewed as the gold standard but are more time and labor intensive than cloned enzyme donor immunoassay (CEDIA). To analyze differences between the 2 assays, we assayed FK506 in 348 split samples simultaneously by both methods. A further 70 samples were stratified by organ transplantation type: cardiac (13%), renal (23%), small bowel (22%), or liver transplantation (42%). Results were analyzed using standard statistical techniques for method comparison. CEDIA overestimated the FK506 value relative to HPLC/MS/MS by more than 20% in 40% of cases (139/348), whereas CEDIA underestimated the FK506 value relative to HPLC/MS/MS by more than 20% in 13.5% of cases, for a total inaccuracy of 53% using a ±20% cutoff. Only 28% of samples (99/348) measured by CEDIA were within 10% of the value obtained by HPLC/MS/MS. Bland-Altman analysis showed a mean bias of 9.5% in favor of CEDIA over HPLC/MS/MS (95% confidence interval, 6.1%-12.9%). Positive bias was greatest for liver transplant and R(2) values were lowest for intestinal transplant patients, indicating that HPLC/MS/MS may be a better option for this pediatric transplant subgroup.