One of the most striking changes in American society in the last forty years has been the decline and delay in marriage. The fraction of young men and women who have never been married increased significantly between 1970 and 2000. Idiosyncratic labor income volatility also increased over the same period. This paper establishes a quantitatively important link between these two facts. Specifically, if marriage involves consumption commitments, then a rise in income volatility results in a delay in marriage. Marriage, however, also allows for diversification of income risk since earnings fluctuations between spouses need not be perfectly correlated. We assess the hypothesis that rising income volatility contributed to the delay in marriage vis-à-vis other explanations in the literature, using an estimated equilibrium search model of the marriage market. We find that the increase in volatility accounts for about 26% of the observed delay in marriage. Thus, we find that the effects of consumption commitments due to increased income volatility outweigh the effects of the insurance gains provided by spouses.