The nicotinic antagonist mecamylamine has been widely shown to cause cognitive impairment. However, these effects are mainly seen with high doses. There have been scattered findings that low doses of mecamylamine can have the opposite effect. This may be due to opposite effects of low doses of mecamylamine. In the current study, an extensive dose-effect function of mecamylamine was characterized in the low-dose range. Adult female Sprague-Dawley rats were trained on a repeated acquisition procedure on an automated 8-arm radial maze. Three of the eight arms were designated as correct for any particular session. Five trials per session were run. The number of errors per trial to find the three correct arms was determined. The rats were trained on the repeated acquisition procedure for at least 18 sessions at which time they showed reliable learning each session. Then, the effect of low doses of mecamylamine between 0 and 1 mg/kg were assessed in a repeated measures counterbalanced design. This dose range of mecamylamine did not affect performance on the first trial when the rats were naïve to the array to be learned. On trials 2-5 a significant (p<.025) quadratic dose-effect function was seen over this dose range. The most substantial effect was seen with 0.125 mg/kg of mecamylamine, which caused a significant (p<.05) improvement relative to the saline control condition. The effect diminished with increasing mecamylamine doses and with the 1 mg/kg dose choice accuracy was back to control levels. This study showed that low doses of mecamylamine can effectively improve learning. A U-shaped dose-effect curve was documented. This suggests possible low-dose nicotinic antagonist lines of treatment for cognitive impairment.