Injection of a dilute solution of formalin into a rat hindpaw produces a biphasic nociceptive response consisting of an early phase during the first 5 min after formalin injection and a later phase starting after 15 min and lasting for 40-50 min. The period between the two phases of nociceptive responding is generally considered to be a phase of inactivity. We compared the nociceptive behaviors produced by a single hindpaw injection of 50 microl of formalin with those produced by two formalin injections given 20 min apart. A single formalin injection at concentrations of either 1 or 2.5%, produced the typical biphasic nociceptive responses. In rats given a second injection of either 1 or 2.5% formalin 20 min after the first, a triphasic response occurred, with a second diminution of nociceptive behavior observed between 10 and 15 min after the second injection. When a second injection of 2.5% formalin was given 5 min after the first, there was no difference from the scores in the group given only one injection. In electrophysiological experiments on single dorsal horn nociceptive neurons, a second injection of 2.5% formalin into the peripheral cutaneous receptive field, 40 min after the first and at the same site of injection as the first formalin injection, depressed neuronal activity for approximately 15-20 min. From the data it can be concluded that the interphase period in the formalin test is due to active inhibition. Furthermore, the inhibition which we are reporting here is independent of the concentration of formalin used, and thus of any so-called inflammatory component. The lack of additive nociceptive effects when the inter-injection interval was only 5 min, suggests that a maximum inhibition was provoked by 2.5% formalin; it can also be concluded that the active inhibition is of overriding importance physiologically, compared with the nociceptive activity. Data from parallel electrophysiological experiments on spinal dorsal horn neurons demonstrated a diminution in excitability after a second formalin injection into the cutaneous receptive field. As these data were obtained from pentobarbital-anesthetized, spinalized rats, the data suggest further that the two excitatory phases and the active inhibition are mediated by spinal mechanisms and that the inhibition is not under regulation of a GABAergic mechanism. The implication of the results is not only further evidence of physiological control mechanisms interacting to regulate pain, but they also indicate the overriding priority of intrinsic inhibitory mechanisms. This, in turn, suggests that the clinical management of pain may be enhanced by efforts to potentiate mechanisms of inhibition.