We agree with Revilla [1] that the resource dispersion hypothesis (RDH) indeed lacks comprehensive evidence in any one specific case; however, this deficiency results from a lack of good tests, rather than from the failure of any tests [2]. Revilla’s claim that ‘we only need evidence against one of its assumptions and/or predictions to invalidate it’, is too sweeping. RDH cannot be rejected just because it does not work everywhere or is difficult to test. Ecological models can only be refined into theories by discovering where they do not work, as well as where they do [3]. The costs of group living that Revilla cites do not detract from the proposition that heterogeneous resources will lower those costs compared with homogenous resources. Revilla sees a ‘conceptual gap’ between the RDH mechanism and a causal means of group formation because he incorrectly assumes that wherever resources are heterogeneous, RDH automatically predicts animals to live in groups. RDH was always a facilitating, rather than a causal, factor leading to group formation 4\ and\ 5. We agree that ‘resource patchiness alone’ is not enough to explain group living (indeed, we have previously outlined why simultaneous dispersal costs are crucial 6\ and\ 7) – RDH is not exclusive of other theories.