Artificial prey, consisting of pieces of pastry placed on coloured cards, were placed on a lawn ; birds, predominantly Blue Tits, were allowed to eat them. When the prey were palatable, the birds ate more from black cards than from red, blue or grey cards. The birds ate significantly more of a "perfect" Batesian mimic than of its model, placed on an identical card, but made distasteful by soaking in a 2 1/2% solution of quinine dihydrochloride. This "discrimination" by the birds probably shows that they can see a slight yellowness in the quinine-soaked pastry. The birds showed no preference between blue, red and yellow prey when the pastry was palatable, but ate significantly more pastry from blue than from red cards when the pastry was distasteful. As this happened equally when the blue or the red prey were in the greater number, it was not due to Mullerian mimicry in the context of the experiment, but probably resulted from Batesian mimicry between the blue distasteful prey and the grey palatable prey which was intended as a control in the experiment. Thus the blue prey, acting as a Batesian model, is shown to suffer increased predation as a result of the predators' confusing it with its mimic.