Male rats were neonatally treated with cholesterol or a substance that blocks the aromatization of testosterone to estradiol (1,4,6-androstatriene-3,17-dione: ATD). At weaning (21 days) they were either housed alone or in small groups (2-3 animals) and tested for partner preference behavior (PPB) in adulthood. Choice was between an estrous female and an active male (Part I) and between an estrous female and an ATD-male (Part II). Tests were carried out in a 3-compartment box. Social isolation did not have major effects on PPB except when sexual interaction with the stimulus animals was prevented (Part I). In this case, isolates (ATD and control) showed higher preference scores (PS) for the estrous female and spent less time in the empty middle compartment. When the choice was between an estrous female and an ATD-male, partner PS decreased in all males, most clearly in ATD-males. The latter animals spent more time with the stimulus ATD-male than they had done in previous PPB tests with the normal stimulus male. In contrast to partner preference behaviors, sexual behavior was clearly affected by social isolation. Isolates (ATD and control) displayed lower frequencies of mounts and intromissions. These effects persisted over consecutive tests. Ejaculation was not affected. In conclusion, the present results confirm earlier findings about the significance of neonatal endocrine conditions for the organization of adult PPB in male rats. The presence or absence of social conspecifics after weaning appears to have little influence on adult PPB.