Hazards on 600 New York farms were audited as part of the NIOSH Farm Family Health and Hazard Surveillance Project. Participants in the audit were recruited by letter and follow up telephone calls to nonresponders to the letter. Letters were sent to a sample of all producers, identified from the New York state agricultural database, with an annual gross income of over 10,000 dollars. Subsequent contacts revealed that not all of the farmers with an initial positive reaction were willing to participate in the hazard audit. All of the surveyors recruited had a farming background and underwent a 1 week training program. Hazard audits were conducted for 4 years in New York state and included visits to farms at appropriate times and assessment of where supplies, equipment and animals were located and general safety on the farm. Farmers were sent a letter 1 week after the farm visit listing observed hazards and explaining why they were hazardous. Positive aspects of farm safety specific to the farm were also mentioned. The authors conclude that an effective surveillance program depends on the personnel employed to carry out surveys (ensuring they have a suitable background and knowledge), developing a trust with those being surveyed (through openness and honesty), timing of contacts, and making nonjudgemental and positive conclusions to the benefit of all parties.