Women’s marital surname change has been discussed as comprising one possible signal of intention to remain married, and may be perceived as such, and valued, by husbands. Here, the practice was investigated as a potential predictor of marital duration among couples who went on to divorce. An archival analysis was based on a search of all available, opposite-sex divorces filed over an 8-month period in a Canadian county. Among couples (n = 107) divorcing, marriages the women in which underwent marital surname change lasted 60% longer, controlling for wife’s age at the time of marriage. When the woman’s marital surname change/retention was used as a regression predictor of number of children of the marriage alongside marriage duration in years, only the latter was predictive. No husband took his wife’s surname. Giving the maternal surname (along with the paternal surname) to children occurred at a negligible frequency. Potential reasons for these findings including costly signaling and, ultimately, paternity uncertainty, as well as possible implications for public policy, are discussed.