The measurement of 16-androstene steroids in submaxillary salivary glands is a potentially good way to estimate boar taint. The objective of this work was to identify factors regulating the concentrations of these steroids in salivary glands. Large White x Pietrain male pigs were divided into four treatments: T-1, intact untreated (n = 18); T-2, intact treated with porcine somatotropin (pST, n = 18); T-3, intact immunized against LHRH (n = 20); and T-4, castrated untreated (n = 19). The pigs were slaughtered at 105 mg of live weight. The mean concentrations of the compounds measured after slaughter for T-1 were as follows: 16-androstene steroid binding protein (pheromaxein) in salivary glands, 2.08 +/- .19 micrograms/g; 16-androstene steroids in salivary glands, 43.3 +/- 2.9 micrograms/g; androstenone in fat, .67 +/- .07 micrograms/g. In all other treatments, the concentrations were reduced (P = .001); T-2 and T-3 had intermediate and T-4 had the least concentrations of the measured compounds. The reduction of 16-androstenes in salivary glands was less than androstenone in fat, probably due to pheromaxein. The overall correlation between androstenone and 16-androstenes was r = .602 (P < .001) and for T-1 was r = .613 (P < .001). The overall correlation between pheromaxein and 16-androstenes was r = .777 (P < .001) and r = .928 (P < .001) for T-1. Only in T-3 was the estimation of the relationship between androstenone and 16-androstenes improved by measuring pheromaxein. These results indicate that pheromaxein is more closely related to the concentration of 16-androstene steroids in salivary glands than in fat. Synthesis of pheromaxein is dependent on mature, functional testes. We conclude that boar taint can be estimated by measuring the concentrations of 16-androstene steroids in salivary glands.