In this paper an attempt is made to give precise expression to the conditions under which a profit maximizing firm with fixed research budget will choose each type of technical change (i.e., "neutral" and "nonneutral"). It was found that the optimal choice depends on the initial technology, relative factor prices, and relative costs of acquiring different types of technical change. The preferred technical change need not be exclusively of one sort (e.g., "neutral" change). Once "neutral" technical change becomes optimal, however, it remains so until there is a change in relative factor prices. On the other hand, adoption of a "biased" technical change may eventually cause "neutral" advance to become desired even in the absence of relative factor price change. Examination of the firm's decision criterion under the assumption that it is a monopsonistic buyer of factors of production, discloses that under identical initial conditions (i.e., relative factor prices and relative costs of alternative forms of technical change) the firm will prefer more "biased" technical change relative to the situation in which it purchases factors competitively. In particular, the firm will, under these condition, seek those "biased" technical changes which economize on the factor whose elasticity of supply is relatively smaller. Finally, it was also discovered that, contrary to previous suppositions, changes in the elasticity of substitution do affect the optimal capital-labor ration for each factor price combination in all cases but one.