Purpose: The objective of this study was to estimate tobacco use prevalence among the Amish in Holmes County, Ohio, using both self-report and a biochemical marker of nicotine exposure. Methods: Amish adults (n = 134) were interviewed as part of a lifestyle study. Self-reported tobacco use was measured using standardized questions, and cotinine was measured from a saliva sample. The prevalence of smoking, total tobacco use, and misclassification were estimated separately by gender, and then compared to 2 non-Amish groups. One group was selected from 2 counties contained within the Holmes County Amish settlement (n = 154) and the other was representative of non-Hispanic whites in the United States (n = 4,099). Findings: No Amish women reported current tobacco use and only 2 reported former use. This was significantly different (P less than 0.0001) from the patterns observed among non-Amish in the settlement counties (15.7%) and US white (23.3%) women. The prevalence of tobacco use among Amish men was 17.6% and this was significantly lower than estimates from non-Amish in the settlement counties (38.8%, P = 0.04) and US white (32.2%, P = 0.005) men. No Amish women and only 2 Amish men underreported tobacco use and misclassification was similar in the comparison groups. Conclusions: Tobacco use is significantly lower among adults in the largest Amish settlement in the world compared to their non-Amish neighbors in Appalachia Ohio and US whites. A strength of this study is that self report was verified with a marker of nicotine, a critical measure to include in any study conducted in a group that stigmatizes tobacco use.