This paper revisits the theory and measurement of monopsony when worker effort responds to wages. We derive a markdown formula incorporating the effort margin and show that ignoring it overstates labor market power. Using estimates of the effort-wage elasticity, we find that implied wage-to-marginal-product gaps in the US fall from roughly 9–48 percent to 3–13 percent. The production approach, however, remains valid with endogenous effort. In a shirking model with imperfect monitoring, stronger monitoring widens markdowns under standard preferences. Consistent with this prediction, firms with more intensive monitoring set wider markdowns in the data.