Helping has many positive consequences for both helpers and recipients. However, in the present research, we considered a possible downside to receiving help: that it signals a deficiency. We investigated whether young children make inferences about intelligence from observing some groups of people receive help and other groups not. In a novel group paradigm, we show that children (4-6 years) think groups that receive help are less smart (n = 44) but not less nice (n = 45). Children also generalized their inferences about relative intelligence to new group members (n = 55; forced-choice-method). These results have implications for understanding how children develop stereotypes about intelligence as well as for educational practices that group children according to their ability.