Trivers' parental investment model states that individuals facing higher levels of parental investment will become increasingly choosy in their choice of mates. For humans, this leads to two predictions. First, both males and females will be choosier in relationships more likely to lead to the production of children. Second, females will be choosier than are males, because their minimum risk of parental investment is higher. Previous studies of human mate choice found support for these predictions, with one curious exception: male choosiness was lower for short-term sexual relationships involving no relationship commitment (one-night stands) than for short-term relationships involving no sexual activity (single dates). Because the risk of parental investment would be higher in a one-night stand, this suggests that true risk of parental investment was not the underlying factor governing choosiness levels, either because study subjects assigned different levels of sexual activity to the relationships than were intended by the investigators of the study or because perceived risk is more important in human mate choice than real risk. To confirm that male/female differences in choosiness criteria exist in humans, and to evaluate the effect that different expected levels of real or perceived parental investment may have on choosiness, we studied mate choosiness in the context of five types of relationships that reflected explicitly defined, increasing levels of risk of parental investment for both males and females. The subjects were 468 undergraduate students, mostly between the ages of 18-24. By using questionnaires, male and female participants rated their minimum requirements in a potential mate for 29 personal characteristics with respect to level of relationship. Our results confirm the major predictions of the parental investment model for humans but suggest that sex differences in choosiness are better explained by perceived rather than real risk of parental investment.