Abstract This article addresses an overview of development and application of an integrated production planning model within the fiberglas industry. The system described represents a second generation in the development cycle of hierarchically based models within this batch production industry. A number of salient issues that distinguish this modeling approach from previous ones are examined. Our contention is that the explicit treatment and integration of demand forecasts within the context of production planning models is a necessary condition for implementation and ongoing use of such systems. Further, due to the nature of practical disaggregate problems, oft cited linear programming (LP) models will be set aside by more flexible goal programming solutions that explicitly recognize the dynamics of variations in product mix, production rates and management objectives. Finally, the development, delivery and maintenance of these planning models will increasingly be performed on microcomputer-based systems and software versus traditional mainframes.