A computer system (MSSMET), using methylene unit retention indices for an off-line reverse library search analysis of selected ion chromatograms from gas-liquid chromatographic mass spectrometric data, has been applied for the qualitative and quantitative determination of daily variations in the excreted levels of urinary steroids of two individuals, using capillary column gas-liquid chromatography. Aliquots of 24 h urine collections and morning spot urine samples were examined. The daily excretion pattern of most of the major steroid metabolites was fairly consistent from day to day (i.e. 3 alpha-hydroxy-5 alpha-androstane-17-one, androsterone; 3 alpha-hydroxy-5 beta-androstane-17-one, etiocheolanolone; 3 alpha, 17 alpha, 21-trihydroxy-5 beta-pregnane-11,20-dione, THE; 3 alpha, 11 beta, 17 alpha, 21-tetrahydroxy-5 beta-pregnane-20-one, THF; 3 alpha, 11 beta, 17 alpha, 21-tetrahydroxy-5 alpha-pregnane-20-one; allo-THF; 3 alpha, 17 alpha, 20 alpha, 21-tetrahydroxy-5 beta-pregnane-11-one, cortolone; 3 alpha, 17 alpha, 20 beta, 21-tetrahydroxy-5 beta-pregnane-11-one, beta-cortolone; 5 beta-pregnane-3 alpha, 11 beta,17 alpha,20 alpha,21-pentol, cortol; 5 beta-pregnane-3 alpha, 11 beta, 17 alpha, 20 beta, 21-pentol, beta-cortol), while certain other steroid metabolites had a less consistent excretion pattern (3 beta-hydroxy-5-androstene-17-one, for example). Advantages and disadvantages of using capillary columns for the automated metabolic profile analysis of urinary steroids by reverse library search of selected mass chromatograms.