We present a theoretical model to determine the optimal management of a size-distributed forest.The decision model is given in form of a distributed optimal control problem that cannot be solved analytically. Thus, the paper presents a numerical technique that allows transforming the original distributed control problem into an ordinary control problem. The method has the advantage that it does not require programming numerical algorithms but rather can be implemented with standard commercial optimization packages such as GAMS. The empirical application of our model for the case of forest management allows determining the selective cutting regime when carbon sequestration is taken into account. © 2008 Springer.