1. The effects of cinnarizine and nifedipine on the nociceptive threshold and opiate antinociception were evaluated by the rat tail-flick test. 2. Male Wistar rats (390-410 g) treated with intraperitoneal (ip) or intrathecal (it) cinnarizine, but not with it nifedipine, displayed a dose-dependent antinociception. The estimated AD50 values of cinnarizine were 3.55 micrograms/kg (confidence limits, 1.99 to 6.32) and 125.9 micrograms/kg (46.1 to 343.7) for the it and ip routes of administration, respectively. The effect of it cinnarizine was reduced by subsequent it administration of calcium chloride (0.1 mumol). 3. The it morphine-induced antinociception was potentiated by the previous it administration of cinnarizine (1.0 microgram/rat). The estimated AD50 of morphine was reduced from 10.4 (6.8 to 16.1) to 4.9 micrograms (3.6 to 6.5) by this dose of cinnarizine. The calculated potency ratio for these values was 2.14 (1.28 to 3.57). A similar potentiation was obtained with it nifedipine, but only when the drug was injected in combination with morphine. 4. It is concluded that the antinociception evoked by a systemically injected calcium channel blocker is dependent on passage of the drug across the blood brain barrier to act, at least in part, at a spinal site of action. 5. The mechanism of the antinociception induced by it injected calcium channel blockers appears to depend on the interaction of the drugs with Ca2+ binding sites in the spinal cord and, probably, on the type of voltage-sensitive calcium channel involved.