Based on published methods for androstenone determination, both an improved enzyme-immunoassay (EIA) and an improved HPLC method were developed and comparatively validated. In contrast to earlier procedures for sample preparation, extraction and solvent distribution were avoided and the separation of androstenone from the contaminating lipids was performed by dissolving the samples in warm methanol followed by cooling down to room temperature (EIA) or down to-20 °C (HPLC). Aliquot portions of the methanol extracts were used for the assays. The EIA was adapted to the contaminating methanol using higher concentrations of gamma-globulins for coating the plates. The determination with HPLC was performed with a fluorescence detector after pre-column derivatization with dansylhydrazine. Comparison of the two systems was performed using spiked reference samples and backfat samples taken from 59 boars. For the EIA, the specificity is based on the antiserum and was found to be fully satisfying for adipose tissue determination. The sensitivity was at 0.04 μg androstenone per g fat. The mean intra- and interassay coefficients of variation were 11.5 and 13.9 % respectively and thus satisfying in regard to the high practicability which allows the determination of about 75 samples per person and hour. The method could also be automatized so that 500 samples per hour are realistic. The HPLC method is less sensitive (0.1 μg/g fat) but the variation is lower. The specificity is based on the chromatographic separation and is sufficient. The method, however, allows the determination of only about 36 samples per day and can be hardly further automatized. The comparative determination in boar samples led to a satisfying agreement as also shown by a correlation coefficient of r = 0.979.