Three quaternary methiodides of nicotine were prepared and tested for antinociceptive activity in the mouse tail-flick, mouse phenylquinone and rat tail-flick tests. Following peripheral administration, all three methiodides were inactive in the mouse and rat tail-flick procedures, whereas nicotine was active in both tests, which suggested that nicotine was acting centrally. Quaternization of nicotine did not eliminate antinociceptive activity as demonstrated by the intraventricular injection of the methiodides in mice. Nicotine pyrrolidine and bis methiodides were somewhat more potent than nicotine, whereas nicotine pyridine methiodide was considerably less potent than nicotine in the tail-flick procedure. Systemically administered nicotine pyrrolidine methiodine was approximately one-third as active as nicotine in the mouse phenylquinone test; nicotine pyridine methiodide and nicotine bis methiodide were 100 and 300 times less active, respectively. Hexamethonium partially blocked nicotine and nicotine pyrrolidine methiodide, whereas mecamylamine blocked nicotine completely but nicotine pyrrolidine methiodide partially. Nicotine may have both central and peripheral actions in the mouse phenylquinone test, whereas nicotine pyrrolidine methiodide may have both nicotine and non-nicotine like antinociceptive activity. The radiolabelled methiodides were synthesized and their disposition in body tissues studied. The methiodides were found to penetrate brain poorly (plasma-to-brain ratios greater than 20). The methiodides were metabolized to nicotine to a small extent. This metabolism occurred to a greater extent in mice than in rats.