Insulin-like growth factor-binding protein extraction protocols were tested for their removing IGFBPs from bovine plasma and bovine granulosa cell culture medium compared with standard acid exclusion chromatography. Traditional extraction methods, acidification, Sep-Pak, ethanol:acetone:acetic acid (EAA) and EAA-cryoprecipitation (EAA-C), failed to remove all the IGFBPs from both granulosa cell culture medium and plasma. However, EAA and EAA-C treatment of plasma samples did give values similar to those obtained by acid exclusion HPLC, when corrected for extraction efficiency. There was an inverse relationship between insulin-like growth factor-I (IGF-I) concentration in plasma samples, as measured using HPLC chromatography, and IGF-I concentration after EAA extraction. Furthermore, the interference caused by residual IGFBPs differed between samples taken from animals given various treatments that altered peripheral IGF-I concentrations. As for plasma samples, EAA was the most effective extraction method for culture media, but residual IGFBPs caused an overestimation of IGF-I concentrations. In culture media, but not plasma, it was possible to block the interference of IGFBPs in the IGF-I assay, in both extracted and non-extracted culture samples, by the addition of excess IGF-II. Using this assay procedure, no IGF-I production by bovine granulosa cells was detected. This was confirmed by HPLC acid chromatography. It is concluded that HPLC extraction is needed for the accurate measurement of peripheral IGF-I concentrations, For granulosa cell culture media it is possible to measure IGF-I concentrations in non-extracted samples if the IGFBPs are blocked by adding IGF-II. Using either this assay, or ai?cr HPLC acid chromatography, no IGF-I was detected in culture media, suggesting that IGF-I is not produced by non-luteinised bovine granulosa cells.