Although personal happiness is highly associated with social relationships, basic questions remain regarding the causal effect of improved social relationships on happiness. The main aim of this study was to test whether emotional and cognitive dimensions of personal happiness can be increased by means of a self-directed relationship enhancement program, which aims to increase relationship-relevant skills such as dyadic coping, communication, and conflict resolution. N = 220 couples were randomly assigned either to an intervention or to a waiting list control group (survey: pre-, post-, 3- and 6-month follow-ups). Multi-group latent change score models reveal that (1) personal happiness can be increased through a couple intervention; (2) change in personal happiness was predicted by an increase in skills and relationship satisfaction through the intervention; (3) the least happy participants benefited most from the intervention. Improving spousal interactions may help people become and stay happier. Participants in this study did not relapse to their set point of personal happiness as could be expected by adaptation theory. Thus, not treating individuals directly, but rather the dynamic system of spousal interactions as a whole, offers a potentially very powerful area for happiness research.