This paper addresses itself to an examination of the desirability of orthogonal expansion in identification and a class of sub-optimal control problems. It is shown that there is little difference between the orthogonal technique and a method based on the direct minimization of the (weighted) integral of error squared or some quadratic performance criterion. It is shown that the orthogonal approach is equivalent to a particular method of matrix inversion. The flexibility and ease of use of both techniques are compared.