In our model where career concerns take the form of promotion, managers use projects and effort to influence the labor market beliefs regarding their ability. We show that managers consider how projects affect the extent to which posterior beliefs can differ from initial beliefs, the precision of posterior beliefs, and equilibrium implementation effort costs. The following results obtain. Although projects differ only in terms of the information they reveal, equilibrium implementation effort costs can reverse project choices. More informative projects can induce lower effort despite the learning effect identified by Holmströmm’s paper (1999). Regardless of the attractiveness of promotion, equilibrium effort is insufficient at either end of the reputation distribution. Finally, good (respectively, bad) reputation managers sometimes choose the more informative (respectively, less informative) project.