The formalin test is commonly used as a model of persistent pain. Besides producing pain behavior, hind paw formalin injection induces the expression of the immediate-early gene, c-fos. A current controversy is whether noxious stimulus-induced Fos protein immunoreactivity can be considered a proxy (biomarker) of nociception in the spinal cord. We investigated this issue by exploiting our recent demonstration of genotype-dependent behavioral differences in response to formalin injection among inbred mouse strains. Accordingly, 6 inbred and 2 outbred strains were administered formalin (5% in 25 microL) into the ventral hind paw, monitored for licking behavior, and then sacrificed at 90 minutes after injection for Fos protein immunocytochemistry. Significant strain differences were observed in both licking behavior and Fos counts in superficial and deep laminae. We observed a significant correlation among strains between licking behavior in the late phase (10 to 60 minutes) of the formalin test and Fos expression in laminae V-VI (but not laminae I-II) of the dorsal horn (r = 0.94). These findings reinforce the use of the Fos technique to study the neuronal processing underlying pain but suggest that Fos labeling reliably reflects tonic pain behavior only in neurons located in the neck of the dorsal horn in mice.