Alternative models of optimal capacity expansion are used to develop a series of hypotheses regarding the behavior of capacity utilization. These hypotheses are then tested at the industry level on a large data sample covering 40 chemical products over a period of approximately 20 years. The results show that capacity utilization was positively influenced by capital intensity and by the trend rate of demand growth. Capacity utilization was negatively influenced by demand variability, the degree of lumpiness of new capacity, the total number of plants in the industry, and the extent of geographic market segmentation. The magnitude of investment scale economies and the extent of multiplant operation appear not to have had any significant effect on capacity utilization.