Gender differences in paid performance have been found in many laboratory-based competitive experiments. They have been attributed to men and women responding differently to psychological pressure. To explore this further, we conducted a laboratory experiment comprising 444 subjects, and measured gender differences in performance in four distinct competitive situations: (i) the standard tournament game where the individual competes with three others and the winner takes all; (ii) an anonymized competition in which an individual competes against an imposed production target and is paid only if he or she exceeds it; (iii) a 'personified' competition where an individual competes against the previous performance of one anonymized person of unknown gender; (iv) a 'gendered' competition where an individual competes against a target based on the previous performance of one anonymized person whose gender is known. Only men responded to pressure differently in each situation; women responded the same to pressure no matter the situation. Moreover, the personified target caused men to increase performance more than under an anonymized target. When the gender of the person associated with the target was revealed, men worked even harder to outperform a woman but strived only to equal the target set by a male.