Discusses the need for convenient, semiformal interactive computer simulation tools and the history of their development. A national economy, a metropolis, the ocean, or the human nervous system are too complex to be analyzed effectively except by simulating them. A computer system designed for simulation studies should provide (1) a means for the nonprogrammer specialist to construct his conceptual model easily and interact with it readily; (2) graphical and nongraphical input; (3) output displays as the solution progresses, enabling the user to intervene as desired; (4) speed commensurate with the user's ability; (5) ease of reentry at any point to alter the model or the conditions and then restart or continue the simulation; (6) convenient information storage, retrieval, and reproduction; and (7) software that handles monotonous details, such as parameter adjustments and warnings of programming problems. Whether present simulation languages can handle biological problems is still uncertain. 10 pp. Ref.