Anabolic-androgenic steroids (AAS) are drugs of abuse. Previous studies have shown that male and female hamsters self-administer testosterone (T) and other AAS, suggesting that androgens are reinforcing in a context where athletic performance is irrelevant. AAS are synthetic derivatives of T, which may be aromatizable to estrogen and/or reducible to dihydrotestosterone (DHT). However, we do not know which metabolites of T are reinforcing. To determine if DHT, estradiol (E(2)), or DHT + E(2) are reinforcing, we tested intracerebroventricular (icv) self-administration in male hamsters. The hypothesis was that androgen reinforcement is sensitive to both androgenic and estrogenic T metabolites. If so, hamsters would self-administer DHT, E(2), and DHT + E(2). Twenty four castrated male hamsters (n = 8/group) received icv cannulas and sc T implants for physiologic androgen replacement. One week later, hamsters self-administered DHT (0.1, 1.0, 2.0 microg/microl), E(2) (0.001, 0.01, 0.02 microg/microl), or DHT + E(2), each for 8 days in increasing concentration (4 h/day). Operant chambers were equipped with an active and inactive nose-poke. At the medium concentration, hamsters self-administered DHT (active nose-poke: 47.9 +/- 13.9 responses/4 h vs. inactive: 18.7 +/- 4.8), E(2) (active: 44.8 +/- 14.9 vs. inactive: 16.6 +/- 2.6), and DHT + E(2) (active: 19.1 +/- 2.4 vs. inactive: 10.4 +/- 2.4, P < 0.05). At the highest concentration, males self-administered DHT (active: 28.3 +/- 7.7 vs. inactive: 15.0 +/- 3.5, P < 0.05) and DHT + E(2) (active: 22.6 +/- 3.8 vs. inactive: 11.6 +/- 2.5, P < 0.05), but not E(2). Hamsters did not self-administer the lowest concentrations of DHT, E(2), or DHT + E(2). These results support our hypothesis that both androgenic and estrogenic T metabolites are reinforcing. Together, they do not exert synergistic effects.