A bioassay technique using isolated guinea-pig ileum was employed to compare the smooth muscle contractor activity of various dusts from mills in which the prevalence of byssinosis was known. The activity of dust from a mill spinning a coarse grade of cotton was several times greater than that in dust from a mill processing a fine grade of cotton. There was a similar order in the difference of the prevalence of byssinosis in these mills. However, the activities of fine cotton, flax, and jute dusts were very similar to each other, in spite of marked differences in the prevalence of byssinosis in these mills. For cotton dust, smooth muscle contractor activity was associated with all particle sizes, although the lowest level of activity was found in the largest sized fraction (less than 2 mm). Activity in the cotton dust extracts was not correlated with nitrogen, carbohydrate, or potassium content. However, about one-fifth of the activity of a cotton dust extract was associated with an insoluble particulate fraction. The possible chemical nature of the water-soluble contractor agent is discussed. It is concluded that, until the role of this agent in the pathogenesis of byssinosis has been established, the bioassay technique cannot be employed as a means of assessing the byssinogenic potential of cotton dust.