A variety of techniques have been applied to monitoring the changes that occur in the resin during composite processing. This report identifies and analyzes those techniques that have the most promise for on-line application, namely: ultrasonic, dielectric, spectroscopic, and optical techniques. All these methods have advantages and disadvantages, and the best technique for a given application will depend strongly on the materials and processes involved. This report discusses these on-line monitoring techniques and compares them in the areas of measurement speed, sampling volume, sensitivity to different resins, effects of fiber type, resistance to the manufacturing environment, interpretation of the data, adaptability for other uses, and temperature capabilities. The most important comparison, however, concerns the research and development work that is needed to achieve the full potential of each technique in applications on the factory floor. Dielectric and ultrasonic methods have an advantage in this connection because commercial equipment designed for process monitoring is already available. The various spectroscopic and optical methods are generally in an earlier stage of development. The great potential inherent in these latter methods, however, makes it highly desirable to actively pursue their developments as well. Although each of the monitoring techniques has its own research and development needs, the most important technical and scientific issues are common to all. From a technological standpoint, the major short term need is to build more rugged and reliable equipment that can withstand the harsh manufacturing environment. Even commercially available systems would benefit from such improvements, particularly in the sensors, electrical connections, and/or electronics, so process monitoring can be used with more confidence on the factory floor. It is also important to develop better systems for operating in electrically noisy environments and with "impure" industrial-grade resins. Solutions to these technical problems are critical to progress in this field, and considerable effort will be required to solve them, but the approach is straightforward. interpretation, and implementation of sensors have not received sufficient attention. Consequently, research and development on sensors is badly needed. Moreover, it could leverage the work performed in projects addressing other aspects of the process control area and thereby play a vital role. The scientific questions outlined above are particularly appropriate since the results will have maximum impact on the development of on-line monitoring techniques. contract programs on this topic. There are major thrusts by the Air Force, NASA, and the National Institute of Standards and Technology. NSF, DARPA, the Office of Naval Research, the Army Research Office, and the Army's Materials Technology Laboratory also have at least some projects on composite process control. Through contracts and subcontracts, many universities are involved in these programs as well. ...