Compared to other leisure time activities, the effect of arts attendance on happiness has received less attention and studies of related topics have produced mixed results, identifying either no effect or very small effects. We investigate this issue using a large (N = 7753) sample from the UK. In contrast to earlier studies, quantile regression is used to allow the relationship between arts attendance and other controls and happiness to vary across different levels of happiness. The relationship found in prior studies is confirmed, but the quantile regression results show that the nature of the relationship varies across the distribution of the happiness variable. A significant relationship is found at the lower quartile, and a moderate relationship (significant at the 10% level) is found at the median level.