A method of deriving rules for optimal fertilizer application when there is carryover is presented. The carryover situations considered are: carryover a deterministic function of fertilizer available in the previous period; carryover a deterministic function of fertilizer applied in more than one previous period; and carryover a stochastic function of fertilizer available in the previous period. The method of derivation is more direct than an inductive one previously presented. The rules are given an economic interpretation, and the gains from using them in practice are considered.
