Mice can learn a food preference from odor cues transmitted on the breath of a conspecific, even if the "demonstrator" is anesthetized. To our knowledge there are no studies examining the effect of anesthetizing the "observer" on development of memory for socially transmitted food preferences (STFP). In Experiment 1 we found that 2-4 month-old F2 C57Bl/6x129sv male and female mice demonstrated a STFP after a 5min exposure to an anesthetized demonstrator mouse when tested 24h later. In Experiment 2, observer mice anesthetized with Sagatal (60 mg/kg) prior to the "social interaction" preferentially avoided the cued food when tested 24h later. This aversion was not due to any overt aversive effects of this dose of Sagatal because mice that ate the food and were then anesthetized, or could only smell the food for 5 min while anesthetized, showed no preference or aversion. In a third experiment we found that the Sagatal-induced aversion was not a general property of anesthesia because there were varied results produced by observer mice treated with anesthetic drugs with different mechanisms of action. Vetalar (200mg/kg) and Rompun (10 mg/kg) treated animals ate similar amounts of cued and non-cued food at test, indicating an absence of learning. Hypnorm (0.5 ml/kg) treated animals showed a preference for the cued food whereas those treated with Hypnovel (2.5 ml/kg) showed an aversion to the cued food. These results show that the food aversion observed with Sagatal is not a general property of anesthetic agents, but appears to be restricted to those acting primarily on the GABAergic system. Thus, we have shown that under certain conditions it is possible for an anesthetized observer mouse to learn a preference or aversion of a socially-linked olfactory cue.