The operational sex ratio (OSR: sexually active males: receptive females) predicts the intensity of competition for mates. It is less clear, however, under what circumstances, the OSR predicts the strength of sexual selection – that is, the extent to which variation in mating success is attributable to traits that increase the bearer's attractiveness and/or fighting ability. To establish causality, experiments that manipulate the OSR are required. Furthermore, if it is possible to control for any OSR-dependent changes in the chosen sex (e.g. changes in male courtship), we can directly test whether the OSR affects the behaviour of the choosing sex (e.g. female choice decisions). We conducted female mate choice experiments in the field using robotic models of male fiddler crabs (Uca mjoebergi). We used a novel design with two females tested sequentially per trial. As in nature, the choice of the first female to mate therefore affected the mates available to the next female. In general, we detected significant sexual selection due to female choice for ‘males’ with larger claws. Importantly, the strength of sexual selection did not vary across five different OSR/density treatments. However, as the OSR decreased (hence the number of available males declined), females chose the ‘males’ with the largest claws available significantly more often than expected by chance. Possible reasons for this mismatch between the expected and observed effects of the OSR on the strength of sexual selection are discussed.