This paper examines whether effects of labor demand shocks on housing prices vary across time and space. Using data on 321 US metropolitan statistical areas, we estimate the medium- and long-run effects of increases in metropolitan statistical area-level employment and total labor income on housing prices. Instrumental variable estimates for different time periods, and also for coastal, non-coastal, large, and small metropolitan statistical areas are obtained using the shift-share instrument. Results suggest that labor demand shocks have positive effects on housing prices. However, these effects appear to vary across time periods and across different types of metropolitan statistical areas.