Changes in the rate of self-directed behavior-a proxy for social tension-offer valuable insights into individuals' experiences of social interactions. Many studies have tested the tension-reduction hypothesis, whereby grooming is expected to reduce social tension in the grooming partners. However, it is still unclear whether responses to grooming are affected by social factors such as an individual's role or social relationship with their partner, and whether responses are similar for females and males. Focussing on same-sex grooming in Yakushima Japanese macaques (Macaca fuscata yakui), we analyzed the effects of two social factors on the change in the rate of self-scratching after receiving grooming: sitting in contact with their partner outside of grooming contexts (as a measure of relationship strength) and rank distance. We collected 10-min postgrooming and matched-control focal samples on recipients of grooming. After controlling for postinhibitory rebound effects, grooming reduced scratching in females (suggesting that grooming was relaxing for the receiver), but not in males. In females, this reduction was associated with increased sitting in contact and increased rank distance: being groomed was more tension-reducing if the grooming partners spent time together outside of grooming contexts, or if their difference in rank was greater and therefore more conspicuous. The reduction effect was mediated by sitting in contact only when females had a higher-ranking-that is, more dominant and potentially aggressive-groomer. Our findings suggest that not all grooming interactions are perceived as equal, which has implications for its use as an index of relationship quality. To our knowledge, this represents the first study of postgrooming social tension in male Japanese macaques and our results highlight the need for caution when generalizing findings from only one sex to the species level.