Mu, delta and kappa opioid receptors in the vertebrate spinal cord mediate the potent antinociceptive effects of opioid agonists administered onto the spinal cord. The present experiments were conducted to determine the effect of unilateral dorsal rhizotomy on mu, delta and kappa spinal opioid binding sites. Measurements of opioid binding were made at 1, 2, 4 or 8 days after rhizotomy and comparisons were made to intact animals. The changes in mu, delta and kappa opioid binding sites were determined by receptor autoradiography using the highly selective radioligands [3H]sufentanil, [3H]DPDPE and [3H]U69593, respectively. Within autoradiograms of each spinal cord, three regions on each side of the spinal cord were targeted for densitometric analysis: laminae I-II (medial), V (lateral) and X. When effects of unilateral rhizotomy within animals were assessed by comparison of the density of binding on the side ipsilateral to the rhizotomy to the contralateral side, decreases in the binding of all three radioligands were observed in laminae I-II on the side of the spinal cord ipsilateral to the rhizotomy at 2-8 days postlesion. A significant reduction in binding was also noted for mu and delta sites in lamina V after 8 days and for delta binding in lamina X at 2 and 4 days on the side ipsilateral to the rhizotomy. However, when densities of binding sites were compared with the corresponding regions in control, it was clear that dorsal rhizotomy resulted in significant changes in opioid binding on both sides of the spinal cord; changes differed for each type of opioid binding site. On the contralateral side of the spinal cord, rhizotomy caused a significant decrease of mu opioid sites 1 day after the lesion and showed partial recovery by day 8. Delta opioid sites were also significantly decreased as early as 1 day postlesion, but did not recover. Kappa opioid sites did not change at 1 day after the rhizotomy but increased on day 2, decreased on day 4 and fully recovered 8 days after rhizotomy. The present results support the hypothesis that a significant proportion of spinal mu, delta and kappa opioid binding sites are present on the central terminations of primary afferents. Finally the present data are the first to report a contralateral effect of the unilateral rhizotomy on spinal opioid binding sites. The contralateral changes in binding were specific to the type of opioid site examined, time after the surgery and region of the spinal cord examined.