The present work assessed the possible participation of K+ channels in the peripheral antinociceptive action of metamizol in the 1% formalin test. Ipsilateral, but not contralateral, local peripheral administration of metamizol produced a dose-dependent antinociception only during the second phase of the formalin test. K+ channel blockers alone did not modify formalin-induced nociceptive behavior. However, local peripheral pretreatment of the paw with charybdotoxin and apamin (large- and small-conductance Ca(2+)-activated K+ channel blockers, respectively), 4-aminopyridine and tetraethylammonium (voltage-dependent K+ channel inhibitors), but not glibenclamide or tolbutamide (ATP-sensitive K+ channel inhibitors), dose-dependently prevented metamizol-induced antinociception. The above results suggest that metamizol could open large- and small-conductance Ca(2+)-activated K+ channels, but not ATP-sensitive K+ channels, in order to produce its peripheral antinociceptive effect in the formalin test. The participation of voltage-dependent K+ channels was also suggested, but since nonselective inhibitors were used, the data await further confirmation.