The European map butterfly (Araschnia levana L.) is a striking example of seasonal plasticity. Individuals of the spring generation are reddish with a fritillary-like colour pattern, whereas the summer generation is black with a white dorsal stripe. Proximate factors explaining the development of the forms are well known, but ultimate explanations have not been tested experimentally yet. The reddish spring form is assumed to have a warning coloration, as found in other nymphalid butterflies that are unpalatable (Aglais urticae). We tested for differential predation by a visually hunting predator (Parus major) in a laboratory experiment using artificial butterflies designed to represent the spring and summer form. Birds were released individually in a flight cage where the alternative forms were presented. Summer forms were more frequently attacked than spring forms, which may point to some aversion against the reddish spring form. But there was also a strong effect of the interaction between seasonal form and type of substrate. Spring forms were much better protected from attacks on the brown substrate of dead leaves compared to the green substrate of nettle leaves. On the latter substrate, latency times before attacking spring forms were on average 2.5 times longer than for summer forms. Experiments with artificial butterflies simplify complex predator–prey interactions because they exclude potential taste or odour effects and they also exclude behavioural responses and interactions of the butterfly. However, our results based on static visual signals provide a promising first step to test the functional significance of this striking seasonal polyphenism.