Male rats were castrated or sham castrated shortly after birth. Castrated males were then injected every other day on days 0-10 with testosterone propionate (TP, 0.5 mg), dihydrotestosterone propionate (DHTP, 0.5 mg) or the oil vehicle (0.05 ml); sham-castrated males received oil injections. In adulthood, when substituted with DHT, DHT + E2, or T (silastic implants), sexual partner preference was measured in an automated open field (AOF), in which wire mesh prevented sexual interaction with incentives, and in a 3-compartment box (3-CB), in which sexual interaction with tethered incentives was possible. Choices were an estrous female and a nonestrous female or an estrous female and a sexually active male. In adulthood, following long-term treatment with DHT or DHT + E2, the males did not show any partner preference when sexual interaction with incentives was prevented. Following sexual experience with an estrous female these males preferred the estrous over the nonestrous female, although this change could also be due to long-term hormone treatment. In the 3-CB, a clearcut preference emerged for the estrous female over the nonestrous conspecific, although the neonatally DHTP- or oil-treated males scored lower than the neonatally TP-treated or control males. Six weeks after removal of the hormone implants, when tested in the 3-CB (estrous female vs. active male), the males showed no partner preference. Unexpectedly the control males showed a low preference for the active male. Three weeks T-treatment made all males show a preference for the estrous female (in 3-CB).(ABSTRACT TRUNCATED AT 250 WORDS)