This paper presents a control system synthesis technique for direct, computer-aided design of control system software with performance improvements in the areas of: sample time determination, control/structural interactions, noise effects (including gusts), dynamic response characteristics, and reduced sensor requirements. The design approach uses a newly developed set of control synthesis computer programs (known as DIGISYN). DIGISYN is based on stochastic control and estimation theory. Control requirements are specified as upper bounds for the state vector error and the desired vehicle response to either control commands or external disturbances. Using a system dynamics model that includes parasitic modes, DIGISYN collectively considers system and sensor noise and external stochastic disturbances (e.g., wind gusts) to determine the maximum permissible sample time, the optimal state estimator, and a set of feedback gains to yield the desired response characteristics. The unique feature of this approach is that sample time is determined by propagating the state covariance matrix until the specified error bounds are exceeded. The rationale for using this method to determine sampling time is that corrective action to reduce the system errors is only applied at the end of each sample period. This paper discusses an application of DIGISYN to pitch-plane control of the Grumman V/STOL Design 607A and verifies performance via a time history simulation. I. Introduction EVOLVING aircraft requirements include extended flight regimes for multimission aircraft and advanced control modes (e.g., automatic landing, speed command, load alleviation) to obtain improved performance. This increasing demand on flight control systems plus the trend toward increased use of computers in aircraft mandates an organized methodology for digital design. In this paper, autopilot is defined as the software link between flight control commands (generated manually or via an automatic guidance law), vehicle mo