Pre-print, Working paper: This paper asks whether a gap in spouses' subjective happiness matters per se, i.e. whether it predicts divorce. We use three large panel surveys to explore this question. Controlling for the life satisfaction levels of spouses, we find that a larger happiness gap, even in the first year of marriage, increases the likelihood of a future separation. This association even holds for couples where both spouses are identified as being better off than in their outside option. We interpret this observation as reflecting a concern for relative utility. To the best of our knowledge, this effect has not been taken into account by any existing economic models of the household. The relationship between happiness gaps and divorce is consistent with the fact that couples who are unable to transfer utility are more at risk than others. It is also possible that assortative mating by happiness baseline level reduces the risk of separation. However, assortative mating cannot entirely explain the finding, as a widening of the happiness gap over time increases the risk of separation. We also uncover an asymmetry in the effect of happiness gaps: couples are more likely to break-up when the difference in life satisfaction is unfavorable to the woman. De facto, divorces appear to be initiated predominantly by women who are less happy than their husband. This asymmetry suggests that the effect of happiness gaps is grounded on motives of relative deprivation, rather than on a preference for equal happiness. The presence of this new argument in spouses' utility is likely to modify their optimal behavior, e.g. in terms of labor supply. It should also be taken into account for public policy measures concerning gender-based labor incentives.