The deficiencies in simulation software's ability to measure the performance of a system's simulation models are described, and it is determined that a unified measurement philosophy is what is lacking in their design. Several symptoms of these deficiencies are suggested. The specialization of simulation software to particular applications or system models such as criminal justice and marketing is believed to be partly symptomatic of the need for such a philosophy. Another reason is found to be the requirement that simulation teams must often append measurement libraries to existing simulation software to augment the neglected measurement approaches available. A consideration of how a unified measurement philosophy might be used to overcome the identified deficiencies defines such a philosophy as a mix of performance measures, measurement strategies, and measurement processes tailored to particular system behaviors, measurement goals, and utilities for information. These characteristics are then developed into an approach for resolving the measurement inadequacies of simulation software based on the types of measures, measurement strategies, and measurement processes which arise in the course of simulation projects. Twenty-six references are provided.