The predictions of optimal parental investment theory were tested by studying nest defence decisions of willow tits,Parus montanus. As predicted, parental defence intensity increased with offspring age, female parents defended early and middle-season broods more vigorously than late ones and the annual proportion of mobbers correlated significantly with annual means of clutch and brood sizes. Against the predictions, defence intensity was not adjusted to the quality (weight, size and condition) or quantity of offspring, which was also confirmed by brood size manipulation experiments. In this respect, willow tits seemed to commit the so-called Concorde fallacy. A suggested explanation for these non-adaptive responses is that a trait to determine the brood value continuously is not needed, since breeding success is high and predictable every year. In addition, in this population parents with reduced and enlarged broods had the same fitness as those with control broods. Therefore, parents behaved optimally when they did not change their defence intensity according to the manipulations. Thus, this lack of response to unnatural variation in a minor component of lifetime reproductive success could be considered a pseudo Concorde fallacy. These results suggest that parents are able to predict the forthcoming breeding success early in the season, and that this prediction holds quite precisely, with parental investment levels being relatively equal regardless of clutch and brood size.