Abstract Nicotine has been found to improve memory performance in a variety of tests, including the radial-arm maze. This improvement, together with the consistent finding of a decline in cortical nicotinic receptor concentration in Alzheimer's patients, has fueled the search for novel nicotinic ligands with therapeutic potential. In the current studies, a series of nicotinic compounds was tested for effects on working memory performance in the radial-arm maze. One of the three compounds tested, DMAE II (dimethylaminoethanol cyclohexyl carboxylate fumurate), produced significant improvements in working memory performance. In the first experiment, this drug produced a biphasic dose-response curve with improved performance at the 20-mg/kg dose but not at 10 or 40 mg/kg. In a second round of DMAE II administration, the same rats showed a significant improvement with the 40-mg/kg dose. In the second experiment, a new set of rats also showed a biphasic dose-response to DMAE II. The 20-mg/ kg dose caused a significant improvement whereas the 40-mg/kg dose did not. Interactions of DMAE II with nicotine and mecamylamine were also studied. Nicotine (0.2 mg/kg) by itself caused a significant improvement in working memory performance. No additive effects of DMAE II with nicotine were seen. In fact, some attenuation of response was seen with the combination. Choice accuracy data for mecamylamine could not be analyzed because of excessive sedation and nonresponding. These studies show that, like nicotine, the nicotinic ligand DMAE II causes an improvement in radial-arm mace choice accuracy. The lack of additivity with nicotine may have been due to the partial agonist effects of DMAE II.