A strategy for design optimization of nonlinearly constrained problems is presented. The strategy combines techniques used in production rule systems with an optimization procedure dealing with local monotonicity and sequential quadratic programming techniques. The rule system is based on observations obtained by applying the optimization procedure to different classes of test problems. The observations made may be incorporated in a rule- based system in such a way that if its premise is true, then the action part of the rule is concluded. This is the first step at developing such a rule-based system for design optimization.