Among animals from the same cage, mice removed last had a higher temperature compared to those removed first. This phenomenon a) persisted 2 and 24 h later; b)was present regardless of the number of the animals (5, 10, 15 and 20) in each cage, c) was independent of whether the number of animals was reduced or maintained constant in the cage and d) could even be observed by reversing the order of removal of the animals from the cage. In addition, the fewer the animals allocated to a cage the greater the percentage of those which became hyperthermic. This rise in rectal temperature of mice removed last was prevented by diazepam (2.5 and 5 mg/kg PO, 30 min), nitrazepam (2 and 4 mg/kg PO, 30 rain) but not by imipramine (15 and 30 mg/kg PO, 60 min) or haloperidol (0.5 and 1 mg/kg PO, 60 rain) and was observed in a greater percentage of mice following subcutaneous yohimbine treatment (2 mg/kg, 60 min). This phenomenon does not seem to depend on physical exercise due to an attempt to escape, since no correlation appears to exist between motor activity (open-field) and rise in rectal temperature. These data would seem to indicate that hyperthermia in the last animals may represent a new tool for studying the neurobiology of anticipatory(?) anxiety.