The optimal amount of reserves that a small bird should carry depends upon a number of factors, including the availability of food and environmental predation risk levels. Theory predicts that, if predation risk increases, then a bird should maintain a lower level of reserves. Previous experiments have given mixed results: some have shown reduced reserves and some, increased reserves. However, the birds in these studies may have been interpreting a staged predation event as a period when they were unable to feed rather than a change in predation risk: theory predicts that, if the food supply within the environment is variable, then reserves should be increased. In the present study, we presented blue tits (Parus caeruleus) with a potential predator and compared this response (which could have been potentially confounded by perceived interruption effects) with a response to an actual interruption in the environment during both long and short daytime lengths. During long (but not short) days, the birds responded in line with theoretical predictions by increasing their reserves in response to interruption and reducing them in response to predation. These results are examined in the light of other experimental manipulations and we discuss how well experimental tests have tested the predictions made by theoretical models.