This is the final version of the article. Available from Wiley via the DOI in this record. ; The cooperative breeding hypothesis (CBH) states that cooperative breeding, a social system in which group members help to rear offspring that are not their own, has important socio-cognitive consequences. Thornton & McAuliffe (2015; henceforth T&M) critiqued this idea on both conceptual and empirical grounds, arguing that there is no reason to predict that cooperative breeding should favour the evolution of enhanced social cognition or larger brains, nor any clear evidence that it does. In response to this critique, Burkart & van Schaik (2016 henceforth B&vS) attempt to clarify the causal logic of the CBH, revisit the data and raise the possibility that the hypothesis may only apply to primates. They concede that cooperative breeding is unlikely to generate selection pressures for enhanced socio-cognitive abilities, but argue instead that the CBH operates purely through cooperative breeding reducing social or energetic constraints. Here, we argue that this revised hypothesis is also untenable because: (1) it cannot explain why resources so released would be allocated to cognitive traits per se rather than any other fitness-related traits, (2) key assumptions are inconsistent with available evidence and (3) ambiguity regarding the predictions leaves it unclear what evidence would be required to falsify it. Ultimately, the absence of any compelling evidence that cooperative breeding is associated with elevated cognitive ability or large brains (indeed data suggest the opposite is true in non-human primates) also casts doubt on the capacity of the CBH to explain variation in cognitive traits. ; A.T. was supported by a BBSRC David Phillips Fellowship (BB/H021817/1) and a grant from the ESRC (ES/M006042/1).