The achromatic number of a graph G is the largest number of colors that can be assigned to the vertices of G so that (i) adjacent vertices are assigned different colors, and (ii) any two different colors are assigned to some pair of adjacent vertices. We study the achromatic number from the point of view of computational complexity. We show that, for each fixed integer n, there is an algorithm which, for an arbitrary graph G, determines in time O(E(G)) whether the achromatic number of G is at least n. In contrast to this, we show that when n is part of the input, the problem is NP-complete even when restricted to bipartite graphs. The complexity of determining the achromatic number of a tree is unknown. We present polynomial algorithms to solve this problem for several classes of trees, and provide upper bounds on the achromatic number of a tree in terms of the maximum degree and the number of edges.