As social robots are increasingly designed with sophisticated simulations of human skills, it becomes essential to understand boundary conditions of people's engagement of them as mindful actors. The present paper reports a comprehensive secondary analysis of six studies (total N = 967) on the relationship between mind perception (evaluation of mental capacities) and mind ascription (explicit assignment of a mind-having status) when people consider a humanoid robot in different scenarios. Results indicate there is a context-independent, moderate link between perceptions of affective mental capacity (i.e., ability to feel emotion) and explicit mind ascription. We further found hints for a weak relationship between perceptions of reality-interaction capacity (i.e., sensory and agentic abilities) and decisions to ascribe mind that may need larger samples in order to be validated. Perceptions of social-moral capacities (i.e., evaluations of people and goals) were not a significant predictor of mind ascription. Overall, these findings highlight the pivotal role played by robots' display of affective engagement with the world for the acceptance of robots as mindful “beings” in human spheres.