This analysis used propensity score matching to construct a comparison sample that is observationally similar at baseline interview to older workers who later experience the onset of a medical condition that limits their ability to work. Using these matched onset and comparison samples, we studied trajectories in earnings and income around onset of the work limitation. Earnings two years after onset for the work-limitation group were 50% lower and poverty rates were nearly double. Income from unemployment insurance, workers' compensation, and retirement and disability benefits offset only a small amount of the earnings declines, resulting in decreased overall household income after onset of the work-limiting condition.