Discussion of the use of general systems theory to improve public policymaking. Different combinations of components from various social macro-systems permit choice, among all potentially effective improvements, of those most feasible under changing conditions. Improvements must reach a critical impact threshold in order to influence the total system. Current improvement efforts (through organizations such as RAND and methods such as PPBS) have limited usefulness because they may be neutralized by the policymaking components and their impact may not reach the critical threshold. Eleven proposals are made to improve present policymaking: (1) explicit strategy decisions, (2) explicit learning feedback, (3) better study of the future, (4) encouragement of creativity and inventions, (5) more analysis, (6) improvement of one-person-centered, high-level decisionmaking, (7) improved training of policy professionals and analysts, (8) improvement of politicians, (9) a distinct "policy science," (10) teaching "good citizenship" to develop individual judgment capacities, and (11) explicit and systematic metapolicymaking.