Two experiments tested whether, similarly to global gender categories, gender subgroups are activated and applied on an automatic basis. In Experiment 1 (N = 90) the female subgroup career woman was activated in a subliminal priming procedure. Experiment 2 (N = 40) contrasted the subliminal priming of career woman with the activation of the global category woman. In both experiments the activated subgroup was applied to a target person in an ostensibly unrelated judgment task: Judgments of the target person were significantly higher on features typical for the subgroup than on untypical features. Activating the global category led to clearly different judgments than when the subgroup was activated. Results are discussed with respect to the use of global and subordinate gender categories in social categorization.