An analysis of the response rate to a health and hazard survey of Ohio grain farmers was performed. The survey was a three phase population based health and hazard survey among cash grain farmers in Ohio conducted as a collaborative effort by Ohio State University, NIOSH, and the State of Ohio. In the first phase of the survey, a postal questionnaire, with telephone followup for the nonrespondents, was sent to nearly 5,000 cash grain farms or farmers throughout the state of Ohio in 1993. The sampling plan, the survey response rate, and the potential for nonresponse bias during phase one of the survey were described. The questionnaire sought information on farm characteristics, pesticide use, work history, tobacco and alcohol use, source of medical care, perceived stress and wellbeing, respiratory health, hearing loss and noise exposure, injuries, knowledge of occupational safety and health topics, chemical exposures, neurotoxicity symptoms, and family history of cancer. The questionnaire response rate varied from 43.6 to 71.4%. The lowest response rate was obtained when subjects who refused to return the questionnaire after the telephone followup (inaccessibles) were included in the calculation. The maximum response rate was obtained when those who returned the questionnaire after the initial mailing and the telephone followup were included in and the inaccessibles were excluded from the calculation. Few differences were seen between the responders and nonresponders that could explain the refusal to complete the questionnaire. Where comparisons were possible, being too busy and health problems were frequently cited reasons for not participating. The authors conclude that the response to this mixed mode survey was low, but not atypical when compared to other surveys of agricultural workers. Methods used for surveying agricultural workers and possible reasons for the low response rates were discussed.