Direct site infusion of drugs into the brain is a powerful tool for examining the function of specific brain regions. While comparing the effects of various drugs injected into distinct regions of the basal forebrain on cognitive and motor endpoints, we observed that the ventral pallidum (VP) appeared to be sensitive to vehicle control infusions when cognitive indices were measured. To characterize this initial observation, the present study examined the effects of vehicle infusions into the VP on performance of a delayed-non-match-to-sample (DNMTS) radial arm maze (RAM) task. A within-subjects design was used. Male Sprague–Dawley rats were trained to perform this task with a 1-h delay imposed between the fourth and fifth arm selection. Following acquisition, animals were implanted with bilateral, indwelling cannulae positioned over the VP. Following surgery, maze performance was reestablished and rats were given one of five intra-VP treatments (two sham and three vehicle infusions) in counterbalanced order. Each rat received one treatment a week, on the third day of five consecutive testing days each week. Vehicle, but not sham, treatments produced deficits on the day of treatment and on two subsequent testing days. These findings demonstrate a persistent sensitivity of the VP to fluid perturbation and, when contrasted with the literature for other basal forebrain regions, it appears that this effect is unique.