Administration of glucose has been shown to enhance performance in a variety of test situations in which memory is impaired by some amnestic treatment. In the present study, the effects of glucose were examined on a working-memory deficit produced when rats performed a delayed matching-to-sample (DMTS) task while being exposed to ambient cold air. In the DMTS task, the rats were required to respond on one of two levers cued by an illuminated light above the lever on the front wall of an operant chamber. Following a variable delay ranging from 1 to 16 sec, both lights were illuminated and the rats were required to correctly respond on the lever previously cued for a food reward. They responded on the back wall lever during the delay interval to prevent position bias. Glucose (10–500 mg/kg) or saline, administered (i.p.) in a mixed sequence, were given 1 h before a 75-min session in which the rats performed the DMTS task (180 trials). During test sessions, the ambient air temperature was either 23°C or 2°C. Administration of glucose during exposure to 23°C did not systematically affect matching accuracy or other performance measures. In the rats pretreated with saline, exposure to 2°C produced a downward shift in the delay gradient in that matching accuracy was impaired at all delay intervals when compared with performance at 23°C. Glucose produced dose-dependent improvement of matching accuracy with short, but not long, interpolated delays. Selective modulation of cold-induced impairment of working memory at the short delays suggests that glucose preferentially enhances stimulus acquisition during exposure to cold stress.