A variety of studies have found that nicotine improves working memory function. However, other studies have either not found improvements or have found nicotine-induced deficits. The demands of the particular memory test may be critical for the expression of the nicotine effects. In several studies, we have found that chronic nicotine administration improves working memory performance in the radial arm maze. Chronic mecamylamine coadministration reversed this effect. The current study was conducted to determine the effects of chronic nicotine and mecamylamine on choice accuracy in a T-maze spatial alternation task. The same dose and duration of nicotine administration that we have previously found to significantly improve choice accuracy in the radial-arm maze was not effective in altering T-maze spatial alternation. The critical difference in task demands may be the presence with T-maze alternation of proactive interference. During a session, a choice alternative repeatedly changes valence from correct to incorrect and back again. In contrast, with the radial-arm maze as run in our studies, in a session the valence of an arm only changes once from correct to incorrect. Previous work with nicotine effects on spatial alternation in an operant task found evidence that nicotine increased the negative effect of proactive interference on performance. In the current study, chronic mecamylamine caused a significant deficit in T-maze spatial alternation. This same dose did not produce a deficit in the radial-arm maze and, in fact, caused an improvement during the first week of administration.