Four methods of dust sampling were compared in three cotton spinning mills in Lancashire. Two of these measured personal breathing zone dust (total inhalable dust and size-selective dust fractions). The other two methods estimated dust levels in the workzone (vertical elutriator and lint-free dust). No fixed convesion factors were identified between the various methods of dust sampling. The ratios between the various fractions varied with the mill and type of process. Generally, however, personal dust was greater than workarea dust, and elutriated levels were lower than other fractions. In the clean mill, however, elutriated dust levels were very similar to lint-free workarea levels and the thoracic fraction of personal dust.