This talk focuses on two recent nursing studies at the University of Michigan Health System (UMHS). The first deals with nurse staffing. UMHS uses three types of nurses: dedicated unit nurses, pool nurses and temp nurses. The key decision is how many unit and pool nurses to hire to minimize the expected cost in the presence of demand uncertainty and nurse absenteeism. The second project deals with assigning roughly 200 surgical nurses to monthly schedules. The problem is inherently infeasible. We solve it as a mixed integer linear programming problem with roughly 100,000 variables and 4,000,000 constraints. Results from both studies are discussed.