Previous research has established that dopamine signals are crucial in orienting behavior to reward. Less is known, however, about the psychopharmacology of task performance under small-reward conditions as compared to large-reward conditions. The current study examined the effects of the noncompetitive N-methyl-D-aspartate (NMDA)-receptor antagonist dizocilpine (MK-801) on reaction time (RT) in a nose-poke task with rats completing an asymmetric reward schedule. In all trials, the rats were required to poke their nose in either the left or the right peripheral hole immediately adjacent to the centre hole when the corresponding light was illuminated. Depending on the stimulus-reward mapping, however, one position was associated with a large reward, while the alternative position was associated with a small reward. Correct performance was required in every trial; if the rat did not make a correct response within 20 s, the trial was aborted, and the same stimulus was presented again on the next trial. In this way, the rat was forced to perform the same visuo-spatial discrimination task under different reward conditions. Reaction times (ms) were faster for large-reward trials than for small-reward trials, replicating previous findings. At a dosage of MK-801 (0.04 mg/kg), there was no significant influence of on RT in large-reward trials. In contrast, the same dosage of MK-801 in small-reward trials produced a decrease in RT as compared to the control condition, implying an improvement of performance. Below 0.04 mg/kg of MK-801, a steady decrease of RT in small-trials was seen as a function of dosage. Above 0.04 mg/kg of MK-801, the majority of rats failed to perform the task at all, whereas the rats that did manage to perform the criterion of 80 correct trials in a session showed no difference in RT between large- and small-reward trials. These data indicate that the systemic administration of a relatively small dosage of MK-801 facilitates performance when reward is small. It is suggested that the facilitation may be due to the reinforcement of mechanisms that work in opposition to response bias. As a corollary, the study provides a useful paradigm to study the voluntary control of unavoidable action.