An attempt was made to clarify the conditions under which humans feel crowded. Four broad types of variables were postulated: physical space, social interaction, duration of exposure, and subject variability.A field survey was taken among residents of a trailer park composed of identical housing units. The subject sample consisted of 58 university students who were asked to rate their living space and their satisfaction with various aspects of trailer living. With total area fixed,the numbers of occupants per unit accounted for greater variance across responses than did social interactions. Time and subject variability accounted for few differences in response. All measures of satisfaction were negatively correlated with increasing degrees of crowding.