Cotton dust and lint are frequently assayed for endotoxin content. A perplexing problem is that all too frequently when different laboratories assay similar or identical samples, major, significant differences exist between the levels of endotoxin reported. This has been observed even between laboratories using the same assay kits and assay protocols. From this we have concluded that some of these differences may result from differences in sample preparation and extraction procedures. Because of these between laboratory differences, interpretation of endotoxin results reported in the literature is difficult and raises the possibility of improper conclusions. Experiments were implemented to determine the effects of extraction variables on endotoxin levels. In particular, the effects of heat, centrifugation to clarify the extract, length and severity of extraction, and the effect of freezing, freeze/thaw cycles, and the materials of the storage container, were examined. Results of these investigations are reported.