We test whether pornography causes divorce. Using state-level panel data on the divorce rate and sales of Playboy magazine, we document a strong cross-sectional and time-series relation between lagged sales of Playboy and the divorce rate. The simple correlation between divorce and sales lagged two years is 44 percent, with a T-statistic of 20. This large correlation is robust to using only the first half of the sample, adjusting for all state-level heterogeneity and for any time trends by including state and year fixed effects, and using an instrumental variable to correct for any possible endogeneity in Playboy sales. Divorce rates are also significantly correlated with sales of Penthouse but they are not correlated with sales of Time magazine. Our overall estimates suggest that pornography probably caused 10 percent of all divorces in the United States in the sixties and seventies.