This paper explores how the use of gender ratios to inform stimulus selection affects the activation of gendered social information. It investigates if stimuli selected this way can activate gender stereotype knowledge and/or conceptual gender knowledge. This was tested through attribute naming (Study 1) and rating (Study 2) tasks, with component and regression analysis allowing for examination of the nature of gender ratios at both attribute and component levels. The results provide rich information about the nature of gender ratio information as a means of stimulus selection, and in doing so support both conceptualisations as long as researchers acknowledge their overlap. The results also indicate that these roles elicited both positive/prescriptive (i.e., the role is appropriate for a given gender) and negative/proscriptive beliefs (i.e., the role is not appropriate for a given gender). These findings hold important implications for future research using gender ratios.