Blood plasma samples collected from 144 healthy volunteers in 16 locations across Canada in 1994 were analysed for ochratoxin A (OTA). The method of analysis included cleanup by C18 solid phase extraction and immunoaffinity columns followed by liquid chromatography (LC) with fluorescence detection, which gave 86.5% (s.d. = 9.7) recovery (n = 31) of OTA (added at 2 ng/ml) with a detection limit of 0.15 ng/ml. The arithmetic mean concentration found in the plasma samples, corrected for volume of anticoagulant added, was 0.88 ng/ml with a standard deviation of 0.35 ng/ml and a range of 0.29-2.37 ng/ml. Confirmation of identity of OTA was by methyl ester formation for 65 samples and by LC-tandem mass spectrometry for 17 samples (some of which were included in pooled samples). Statistical analysis, by ANOVA of the log OTA plasma concentrations, showed a highly significant effect due to location in Canada (p = < 0.0001) but no effect due to age, sex or blood group of donors. The highest mean concentration was found in Winnipeg, significantly different (p = 0.05) by the Student-Newman-Keuls multiple range test from the lowest levels found in Toronto, Vancouver and Saint John.