An enzyme-linked immunosorbent assay (ELISA) measuring serum infliximab concentrations in treated patients was developed. Microtiter plates were sensitized with tumor necrosis factor alpha (TNF-alpha) and saturated with phosphate-buffered saline (PBS) containing 1% bovine serum albumin (BSA). Samples diluted 1:100 in PBS-1% BSA were added and bound infliximab was detected using peroxidase-conjugated goat anti-human immunoglobulin G specific for Fc fragment (HRP-anti hIgG). Reading was performed using an ELISA plate reader. The limit of detection, calculated by assaying 10 replicates of a drug-free serum sample or blank sample and defined as the lowest concentration distinguishable from zero at 2 standard deviations, was 0.014 microg/mL. Each quality control sample was tested on 7 occasions on 1 day and on 5 separate days. The intraday precision indices of the method were (percent coefficients of variation, CV%) 11.7%, 6.2%, and 6.9% for 0.04 microg/mL, 2 microg/mL, and 4.5 microg/mL, respectively. The corresponding bias measures (percent deviation) were -5.5%, -1.9%, and -7.9%, respectively. The between-days precision was 9.8%, 5.3%, and 5.3% for 0.04 microg/mL, 2 microg/mL, and 4.5 microg/mL, respectively. The corresponding bias were +0.3%, -0.3%, and -7.8%, respectively. Lower limit of quantitation and upper limit of quantitation were 0.04 microg/mL and 4.5 microg/mL, respectively. Trough serum concentrations of infliximab were measured in 6 adult patients with various diseases and in 5 pediatric patients with Crohn's disease. For the latter group, samples drawn 1 hour after the end of the infusion and repeated measurements also were available. Data were described using a 1-compartment population pharmacokinetic model. Terminal elimination half-life was 10.9 days. This method is rapid, accurate, and reproducible, and may be useful in therapeutic drug monitoring of infliximab.