In a continuation of work on the byssinosis problem, cotton fiber was field-weathered before harvest for varying periods on plants at locations across the U.S. Cotton Belt and counts were made on the fiber for total, gram-negative, and coliform bacteria. At Raleigh, NC, Auburn, AL, and Stoneville, MS, bacterial counts increased greatly during the first week of weathering, decreased during a dry period in mid-to-late September, then increased again later. At Florence, SC, the fiber was exposed to repeated rains and exhibited a generally increasing level. Counts on fiber from Shafter, CA were very low, the data thus resembling earlier data from the same location. Counts on fiber from insect-damaged bolls from Brownsville, TX generally displayed variable but often exceedingly high bacterial levels.