We analyzed the mesothelioma mortality in cohorts of workers exposed to crocidolite, amosite, and chrysotile to estimate asbestos fiber potency for mesothelioma, using the method of Hodgson and Darnton (2000). We relied on the original 17 cohort studies in their analysis, along with 3 updates of those studies and 3 new asbestos cohort studies published since 2000. We extended the analyses to examine the mesothelioma potency of tremolite in vermiculite from Libby, Montana, and for non-asbestiform elongate mineral particles (EMPs) in taconite iron ore, talc, and South Dakota gold mining. Mesothelioma potency (R_Meso) was calculated as the percent of all expected deaths that were due to mesothelioma per fiber/cc-year of exposure.The R_Meso was 0.0012 for chrysotile, 0.099 for amosite, and 0.451 for crocidolite: thus, the relative potency of chrysotile:amosite:crocidolite was 1:83:376, which was not appreciably different from the estimates by Hodgson and Darnton in 2000. The R_Meso for taconite mining fibers was 0.069 which was slightly smaller than that for amosite. The R_Meso for Libby fibers was 0.028 which was greater than that for chrysotile and less than that for amosite. Talc and gold mining EMPs were non-potent for mesothelioma. Although there are a number of methods for estimating fiber potency of asbestos and non-asbestiform EMPs, the method of Hodgson and Darnton provides a uniform method by which fiber potency can be compared across many fiber types. Our estimates of R_Meso provide a useful addition to our knowledge of mesothelioma potency for different asbestos and non-asbestiform EMP fibers.