The present study investigated the involvement of corticotrophin-releasing factor (CRF) in the thermogenic and anorexic actions of serotonin (5-HT) in the rat. Serotonergic compounds and CRF antibody were injected directly into the third ventricle of conscious, male Sprague-Dawley rats. Thermogenesis was measured as changes in whole body oxygen consumption by indirect calorimetry. Central injections of 5-HT (0.5-50-mu-g, i.c.v.) significantly increased resting oxygen consumption (V(O2); maximum 12.5% increase), without obvious effects on behaviour. Similar increases in V(O2) (12-17%) were observed following central injections of the 5-HT precursors, tryptophan (14-mu-g, i.c.v.) or 5-hydroxytryptophan (5-HTP, 20-mu-g, i.c.v.), and peripheral (10 mg/kg, i.p.) or central (30-mu-g, i.c.v.) injections of the 5-HT reuptake inhibitor, DL-fenfluramine. Administration of a polyclonal CRF antibody (3-mu-l, i.c.v.) 10 min prior to serotonergic compounds, significantly reduced (77-106%) the increases in V(O2) observed in response to central injections of 5-HTP (20-mu-g), 5-HT (50-mu-g) or peripheral injections of fenfluramine, but not those observed in response to either 30-mu-g fenfluramine (i.c.v.) or 20-mu-g 5-HT. Voluntary food intake was measured for 6 h in rats following 16 h starvation. Six-hour food intake was significantly reduced (30-60%) in rats given central injections of 5-HT or 5-HTP, and central or peripheral injections of fenfluramine. Pretreatment with CRF antibody significantly attenuated (30-70%) the reductions in food intake observed in response to central injection of 5-HT (40-mu-g), 5-HTP (40-mu-g) and fenfluramine (30 and 50-mu-g), but not those observed after either 20-mu-g 5-HTP, 20-mu-g 5-HT or peripheral injection of fenfluramine (10 mg/kg). The central thermogenic and anorexic actions of serotonin and related compounds appear to be mediated, at least in part, by CRF.