The present study compares the potencies of D-Ala2-D-Leu5-enkephalin (DADL) and beta-endorphin (beta-EP) injection intrathecally and i.c.v. on the inhibition of tail-flick and ice water-induced shaking responses in pentobarbital-anesthetized rats. Peptides were injected stereotaxically into the third ventricle or into the spinal subarachnoid space. The tail-flick response was measured 10 or 20 min after the injection of DADL or beta-EP, respectively, after which the rats were immersed for 5 min in ice water and the number of shakes was counted. Intrathecal DADL and beta-EP were equipotent in inhibiting both the tail-flick and shaking responses. However, treatment with naloxone (2 mg/kg i.p.) shifted the dose-response curve for both responses to the right more for the beta-EP than for DADL. Unlike the equipotency intrathecally, DADL by i.c.v. injection was less potent than beta-EP in inhibiting both the tail-flick and shaking responses. Previous results in the unanesthetized rat had also shown that i.c.v. DADL was less potent than beta-EP in inhibiting the tail-flick response. Thus, the spinal cord is sensitive to both DADL and beta-EP, whereas the supraspinal area is more sensitive to beta-EP. Also, because the potency of i.c.v. beta-EP for inhibiting tail-flick but not shaking was substantially reduced in anesthetized compared to unanesthetized rats, the neuronal substrates involved in inhibition of shaking are different from those of inhibition of the tail-flick. The differential sensitivities to i.c.v. vs. intrathecal administration and the unequal responses to the antagonistic action of naloxone indicate that DADL and beta-EP produce their action through different opioid receptors.