This paper studies the effects of unions in private-sector nursing homes on a broad range of labor, firm, and consumer outcomes. We link national data on nursing home characteristics from the Centers for Medicare and Medicaid Services to records on establishment-level unionization from federal labor agencies, and employ a regression discontinuity design to identify union effects by contrasting outcomes in nursing homes where unions closely won representation elections to outcomes in facilities where unions closely lost such elections. After showing that these two sets of homes are similar leading up to the election, we estimate union effects on staffing levels, care quality, and other outcomes. We find negative effects of unions on staffing levels and no decline in care quality, suggesting positive productivity effects. Consistent with these results, supplementary analysis shows significant increases in wages for some classes of nursing labor. Some evidence suggests that nursing homes in local product markets that were less competitive and had lower union density at the time of election experienced stronger union employment effects. We find no impact of unionization on facility survival. By combining credible identification of union effects, a comprehensive set of outcomes over time with measures of market-level characteristics, this study generates some of the best evidence available on many controversial questions in the economics of unions. Furthermore, it generates evidence from the service sector, which has grown in importance and where evidence on these questions has been thin.