The effect of intracerebral injection of TRH and several biologically stable TRH analogues in the pentobarbitone anaesthetized rat was examined. Bilateral injection of TRH (5.0 micrograms total dose) and the analogues RX 77368 (0.01-1.0 microgram), CG 3509 (0.1-1.0 microgram), DN-1417 (1.0 microgram) and MK-771 (1.0 microgram) into the nucleus accumbens reduced the pentobarbitone-induced sleeping time. The TRH metabolite DKP (5 micrograms) had no effect on the sleeping time following intra-accumbens injection. Intra-septal injection of TRH (1.0-5.0 micrograms), RX 77368 (0.1-1.0 microgram) and CG 3509 (0.1-1.0 microgram) also reversed the pentobarbitone-induced sleeping time. In contrast, TRH (5 micrograms) injected into the striatum had no effect on the pentobarbitone-induced sleeping time, and CG 3509 (0.1 microgram) and RX 77368 (0.1 microgram) had weaker effects following intrastriatal injection compared to injection of these analogues into the nucleus accumbens and septum. Measurements of core temperature and respiration rate in rats following intra-accumbens or septal injection of TRH, CG 3509 and RX 77368 showed these peptides to reverse pentobarbitone-induced hypothermia and stimulate respiration rate. However, while intrastriatal injections of CG 3509 and RX 77368 caused an increase in respiration rate they had no effect on core temperature. These results suggest a close association between peptide-induced respiratory stimulation and reversal of pentobarbitone-induced anaesthesia. Since intra-accumbens and septal injection of dopamine (20-100 micrograms) failed to reverse anaesthesia, it is unlikely that the peptide-induced responses are mediated via dopamine release.