We present a model that predicts the maximum axonal length from the apparent velocity of slow axonal transport and cytoskeletal protein half-life. The model assumes that in mature axons the apparent velocity of slow transport varies with position, but that the density of cytoskeletal proteins and protein degradation are uniform. The model predicts that the apparent transport velocity of cytoskeletal proteins if highest near the cell body and decreases linearly along the axon, and that when axons branch the apparent velocity of transport decreases across the branch point. The predictions of this model are shown to be consistent with experiments. These results explain the variation in these fundamental metabolic parameters in different axons and species.