The Abbott TDx fluorescence polarization immunoassay (FPIA) system was evaluated and compared with well-established enzyme multiplied immunoassay technique (EMIT) and radioimmunoassay (RIA) methods utilizing five high-volume drug assays including theophylline, gentamicin, phenytoin, phenobarbital, and digoxin. These drug assays were evaluated for precision, calibration stability, specificity, and accuracy. Within-run precision studies utilizing control samples (n = 20) in the subtherapeutic, therapeutic, and toxic ranges resulted in coefficients of variation (CV) of less than 4.0% for the theophylline, gentamicin, phenytoin, and phenobarbital assays and of less than 9.5% for the digoxin assay. Between-run precision studies based on an initial TDx calibration curve over a 2-3 week period yielded CVs of less than 8% for all five drug assays. Cross-reactivity of the FPIA gentamicin assay with concurrently used aminoglycosides such as tobramycin and amikacin was less than 0.1%, and interference due to hemolysis and lipemia was negligible. Highly icteric specimens resulted in clinically significant decreases in theophylline and phenytoin concentrations, but this problem can be corrected by subtraction of blank intensity values. Comparison of the FPIA method with the EMIT and RIA methods indicated an extremely good analytical correlation (r greater than 0.97) for all five comparisons. The Abbott TDx FPIA system offers significant advantages in calibration and reagent stability, and greater sensitivity in the low drug concentration ranges while maintaining accuracy and precision comparable with those of established EMIT and RIA procedures.