"It’s the same as pre-order, except that we visit the root after the children instead of before. Still, despite its similarity, this has always been the trickiest one for me. Everything seems postponed, and you have to remember what order to do it in later. For our sample tree, the first node visited turns out to be I. This is because we have to postpone visiting G until we finish its left (and right) subtree; then we postpone K until we finish its left (and right) subtree; postpone D until we’re done with O’s subtree, and postpone O until we do I. Then finally, the thing begins to unwind...all the way back up to K. But we can’t actually visit K itself yet, because we have to do its right subtree. This results in C, E, and M, in that order. Then we can do K, but we still can’t do G because we have its whole right subtree’s world to contend with. The entire order ends up being: I, O, D, C, E, M, K, A, F, L, N, B, H, and finally G. Note that this is not remotely the reverse of the pre-order visitation, as you might expect. G is last instead of first, but the rest is all jumbled up. The order of node visitation in post-order traversal is shown in Figure"