Problem-solving skills are predicted by technical innovations in the wild and brain size in passerines.

Behavioural innovations can provide key advantages for animals in the wild, especially when ecological conditions change rapidly and unexpectedly. Innovation rates can be compared across taxa by compiling field reports of novel behaviours. Large-scale analyses have shown that innovativeness reduces extinction risk, increases colonization success and is associated with increased brain size and pallial neuron numbers. However, appropriate laboratory measurements of innovativeness, necessary to conduct targeted experimental studies, have not been clearly established, despite decades of speculation on the most suitable assay. Here we implemented a battery of cognitive tasks on 203 birds of 15 passerine species and tested for relationships at the interspecific and intraspecific levels with ecological metrics of innovation and brain size. We found that species better at solving extractive foraging problems had higher technical innovation rates in the wild and larger brains. By contrast, performance on other cognitive tasks often subsumed under the term behavioural flexibility, namely, associative and reversal learning, as well as self-control, were not related to problem-solving, innovation in the wild or brain size. Our study yields robust support for problem-solving as an accurate experimental proxy of innovation and suggests that novel motor solutions are more important than self-control or learning of modified cues in generating technical innovations in the wild.