Scent marking is often assumed to be a secondary sexual trait that increases males' mating and reproductive success, although direct evidence for this hypothesis is lacking. We conducted a study with wild-derived house mice, Mus musculus musculus, to test whether scent marking increases males' reproductive success when females can freely choose between two territorial males. We also experimentally manipulated males' competitive scent marking by exchanging scent-marked tiles between the neighbouring males' territories (intrusion treatment) or relocating males' tiles within their own territory (control). Experimental animals were tested twice and we examined whether individual males were consistent in their marking. We found that males marked more in the intrusion treatment than controls and more at shared territorial borders than elsewhere. We found high day-to-day variation in most individuals' scent marking, and yet the sum of individuals' scent marking was consistent over time and across different social conditions. Genetic paternity analyses revealed that males' scent marking significantly increased their reproductive success in both the intrusion treatment and the controls. Surprisingly, however, female social preference was not positively correlated with male scent marking. These results provide direct evidence that scent marking increases males' reproductive success when females can choose their mates, even though it did not increase females' social preferences.