Here we show that gender identification of male (but not female) heterosexual, right-handed dancers correlates with physical strength (measured via handgrip strength) after controlling for the effect of body-mass-index on strength. Using optical motion capture technology, we collected the dance movements of men and women for subsequent animations of uniform shape- and texture-standardized virtual characters (avatars). Short video clips (15. s) of these movements were presented to male and female adults and children, who were asked to identify the gender of the avatar. Gender identification performance was significantly higher than chance for both adults and children. Among adults (but not among children) the avatars of male dancers who were physically stronger were perceived as males significantly more often than were the avatars of male dancers who were physically weaker. There was no relationship between strength and gender identification for female dancers. We conclude that physical strength affects gender identification from human dance movements at least for male dancers, and that pre-pubertal children might not be sensitive to strength cues in dance movements.