From a cognitive perspective, mate choice consists of three interrelated subtasks: (a) identifying features that are relevant for choosing a good mate, (b) integrating the features into a single attractiveness judgment, and (c) searching for possible partners. Computer simulations show that simple cognitive mechanisms can solve these tasks quickly and successfully. For instance, in a sequential search process such mechanisms lead to good mate choices by generating an aspiration level based on the quality of a small number of previously encountered candidates and on the feedback obtained on one’s own mate value. This aspiration level is then used to gauge the quality of future candidates; the first candidate subsequently encountered that exceeds this aspiration level is chosen. An empirical study of “fastdating” provides support for some of the predictions of the simulations and interestingly, also shows that stated preferences often do not match the features that underlie the actual mate choices.