A new and more profitable food type was introduced to three groups of Mongolian gerbils (Meriones unguiculatus). When first presented with the new food one group was sated, another was hungry and the third group was hungry and expected a food shortage. In addition to the new food all groups had access to familiar food during the experimental sessions. The group that expected a food shortage had a lower intake of new food than either of the other groups, both in absolute and relative numbers. Both the sated and the hungry group ate the same absolute amount of new food during the first presentation, but since the sated animals had a lower total intake, the new food represented a higher proportion of the intake for these animals. The animals were then tested during the subsequent days (all groups now hungry). The group that had been sated showed a stronger preference for the new and profitable food than the hungry group, which in turn showed a stronger preference than the expected food shortage group. This means that in a semi-natural situation a strong need may actually decrease the rate of learning. The results can be explained in terms of risk prone/risk aversive behaviour along the lines of previous studies of risk sensitivity in optimal foraging. In an additional experiment it was shown that sated individuals will decrease their total intake of food when presented with both new and familiar food, in comparison with when only familiar food is present. This may have the effect of enhancing learning about the new food.