In the context of an automated psychophysical laboratory, the adequacies and inadequacies of a small computer are discussed. Attention is given to the rationale for choosing the implementation in particular a dedicated small system as opposed to time-shared use of a larger system. The variables discussed include cost, flexibility, language level, and storage capabilities in addition to reliability. A compromise position that appears most viable for the future is also suggested. © 1969 Psychonomic Society, Inc.