Central to sexual selection theory is the question of when individuals should compete for mates. Theory predicts that the sex ratio of ready-to-mate individuals (operational sex ratio; OSR) affects male and female mating competition. In accordance with this, the strength of mating competition, measured by agonistic behaviors and courtship displays, has been found to co-vary with the OSR in field populations of several species. However, laboratory experiments have often produced results that seemingly contradict OSR theory, especially for courtship behavior. We argue that this may be because experiments typically measure frequencies of competitive behaviors. Frequencies of courtship and agonistic behavior are not only affected by the level of mating competition, but also by the number of potential mates or competitors encountered. In contrast, the propensity to behave competitively at a given encounter represents a behavioral response, and thus directly reflects mating competition. We show in 2 simple models that 1) courtship frequency can be expected to respond differently from courtship propensity to changes in OSR and 2) an increase in frequency of agonistic behaviors could occur even if the propensity is not affected by the OSR. In a meta-analysis of studies on courtship competition, we show that frequency measures produced largely opposite results to propensity measures, as predicted by our model. Moreover, courtship propensity increased when the OSR became more biased toward competitors. This presents strong evidence that the OSR affects competition, in the form of courtship, as predicted by OSR theory.