The study of human mate preferences has flourished over recent years, in large part thanks to the impact of evolutionary considerations. However, this work has been criticised for overly relying on self-report data (Cooper & Sheldon, 2002). A recent study by Eastwick and Finkel (2008) sought to address this issue by investigating actual mate selection behaviour in a speed-dating setting, but they found, surprisingly, that although participants’ reported mate preferences replicated previous research, participants’ actual choices did not. In this thesis I describe two studies that aimed to partially replicate and explain these findings by asking participants to imagine themselves in a hypothetical internet dating scenario. In Study 1 (N = 104), instructions were designed to manipulate both the consequence of obtaining a match with a target and the control participants had over whether they obtained a match. Contrary to my hypotheses, neither of these manipulations influenced hypothetical mate choices. However, as predicted, correlational indices of preference for attractive targets were related in a curvilinear fashion to number of targets picked, suggesting that such indices may suffer from restriction of range problems. Moreover, across experimental cells participants’ choices were broadly consistent with evolutionary considerations, and sex differences in preference for attractive targets were mediated by participants’ interest in short-term relationships. These latter two findings were replicated in Study 2 (N = 71) using a forced choice paradigm along with alternate measures of preferred target. These two studies suggest that individuals make mate choices that are predictably different across sex and held ideal standards. The results were also broadly consistent with evolutionary theory.