Abstract: Chimpanzees undoubtedly form concepts related to the statistical regularities in behavior. But do they also construe such abstractions in terms of mental states – that is, do they possess a [`]theory of mind'? Although both anecdotal and experimental data have been marshaled to support this idea, we show that no explanatory power or economy of expression is gained by such an assumption. We suggest that additional experiments will be unhelpful as long as they continue to rely upon determining whether subjects interpret behavioral invariances in terms of mental states. We propose a paradigm shift to overcome this limitation.