We identify, quantify, and explain the impact of incentive-induced early retirement (ER) of husbands on their wives' probability to retire within 1 year, using administrative data from the Netherlands. Our identification strategy is based on a policy intervention by which targeted individuals working at the central government level became unexpectedly and temporarily eligible for very generous ER benefits. This retirement window of opportunity implied for interested workers that transitions from the current job into full retirement had to be effected swiftly and irreversibly. We find that induced ER of husbands increased their wives' probability to retire by 10 percentage points. This is a strong and robust local average treatment effect. Partly, the effect runs through wives at ages when they may have been eligible for ER programs themselves. (JEL C26, J26, J120, J140).