What a difference a day makes! Using data from government registries on all the weddings that occurred in the Netherlands from 1999-2013, we find that weddings that occurred on Valentine's Day, dates with the same numbers for the day, month and year like 9 September 1999 (9.9.99), and dates with numbers for the day, month, and year in an ascending sequence like 1 February 2003 (1.2.03) were both exceptionally popular and also much more likely to end in divorce compared to weddings that occurred on ordinary dates.How popular? There were nearly three times as many weddings in the Netherlands on a given Valentine's Day than on other comparable days in February, and there were six times as many weddings on a given same-number day as on a comparable ordinary day.However, the special-day marriages proved to be fragile. By what would have been their ninth anniversaries, 21 percent of Valentine's Day marriages and 19 percent of same-number-date marriages were estimated to fail, but only 16 percent of ordinary-date marriages were estimated to fail. In relative terms, Valentine's Day marriages were about one-third more likely to fail by their ninth anniversaries and same-number-date marriages, about one-quarter more likely to fail than ordinary-date marriages. We continue to find large differences in divorce risks even after we use statistical methods to account for characteristics that might affect both the choice to marry on a special date and marriage vulnerability.Although an analysis of quirky wedding dates is unusual, our results match predictions from wellknown family-relations theories of how commitments deepen in couple relationships, especially the "sliding vs. deciding" theory. From this theory, "deciding" couples choose to marry because of what they learn about each other and about their compatibility over the course of their relationships. In contrast, "sliding" couples slip into marriage because constraints to exiting the relationship, such as the costs of moving or of separating bank accounts, gradually build and make marriage seem worthwhile, even if the couple is less compatible. The timing of "deciding" couples' marriages will depend on how information about their relationships develop, while the timing of "sliding" couples' marriages may be more influenced by external events, like the opportunity to marry on a romantically or numerically special day. Consistent with this theory, we also find that special-day couples are less alike and less well-matched at the start of their marriages compared to ordinary-day couples.