This study evaluated the effects of intrathecally coadministered morphine and the N-methyl-D-aspartate (NMDA) antagonist (+)5-methyl-10, 11-dihydro-5H-dibenzocyclohepten-5,10-imine maleate (MK-801) on the thermally evoked hindpaw withdrawal latency (PWL) in rats with one paw (ipsilateral) rendered hyperesthetic by the unilateral application of loose ligatures to the sciatic nerve (DELTA-PWL (+/- S.D.) = PWL(hyperesthetic paw) - PWL(normal paw) = 3.1 +/- 1.2 s). Intrathecal morphine produced a dose-dependent (0.1-10-mu-g; P < 0.0001) elevation in the thermal response latency of both the contralateral (normal) and ipsilateral (hyperesthetic) paw. DELTA-PWL did not vary with morphine, indicating that the dose-response curves were parallel but shifted to the right for the hyperesthetic paw. For the normal paw, MK-801 (10-mu-g) was without effect upon the response latency; whereas, the response latency of the hyperesthetic paw was elevated to the same as the normal paw, i.e. the hyperesthesia was selectively abolished (DELTA-PWL (+/- S.D.) = -0.067 +/- 2.73). Co-administration of MK-801 with morphine did not alter the effects of morphine in the normal paw, but reduced the DELTA-PWL for each dose of morphine. These results suggest that NMDA antagonism (1) does not alter the thermal sensitivity in the normal paw, (2) selectively abolishes the hypersensitivity of the hyperesthetic paw and (3) has a simple additive interaction with the antinociceptive effects of morphine in the hyperesthetic paw.