Love is commonly hypothesized to function as an evolved commitment device, disincentivizing the pursuit of romantic alternatives and signaling this motivational shift to a partner. Here, we test this possibility against a novel signaling-to-alternatives account, in which love instead operates by dissuading alternatives from pursuing oneself. Overall, we find stronger support for the latter account. In Studies 1 and 2, we find that partner quality relative to alternatives positively predicts feelings of love, and love fails to mitigate the negative effects of desirable alternatives on relationship satisfaction—contradicting the classic commitment device account. In Study 3, using a longitudinal design, we replicate these effects and find that changes in partner quality relative to alternatives predict changes in love over time. In Study 4, we replicate the relationship between love and relative partner quality across 44 countries. In Study 5, we find a nearly one-to-one correspondence between the extent to which partner-directed actions are diagnostic of love and reductions in romantic alternatives' attraction to the actor. These results suggest that love may not act as a commitment device in the classic sense by disincentivizing the pursuit of alternatives but by disincentivizing alternatives from pursuing oneself.