We explore management science options for scheduling the placement of fuels reductions. First, we look at approaches for creating and maintaining a prespecified set of forest conditions that are deemed desirable from a fuels management perspective. This approach is difficult because the pre-specified forest is hard to determine and because it is an any-aged forest management problem that is intrinsically nonlinear. Second, we look at capturing the spatial relationships suggested by fire behavior in long-term fuels management. This approach is difficult because fire origins and behavior can be quite random and unpredictable. It is necessary to accept a particular fire event (analogous to the 500-year flood in flood-control planning) as the target for fuels management. Overall, we conclude that long-term fuels management presents a formidable problem for management scientists. Numerically intense stochastic programming methods, such as Monte Carlo repetitions with a fire simulator, may be the most promising approach.