We compared a functional (amidolytic) and an enzyme-linked immunosorbent assay (ELISA) method for determining aprotinin concentration in 82 plasma samples obtained from patients undergoing cardiac surgery with aprotinin therapy. There was good correlation between methods (r = 0.87); however, aprotinin measurements by chromogenic assay were significantly higher than by ELISA [234 +/- 104 kallikrein inhibitory units (KIU)/ml versus 155 +/- 88 KIU/ml; P = 0.0001]. This appeared to be attributable to differences in the potency of the material used to standardize the assays. When results were corrected to allow for potency of the standard, there was no significant difference between chromogenic and ELISA methods (234 +/- 104 KIU/ml versus 240 +/- 137 KIU/ ml), although the ELISA results tended to be higher in some samples. These data suggest that aprotinin concentrations measured by these methods cannot be used interchangeably, and care must be taken when interpreting data from studies measuring aprotinin.