To better estimate rates of certain benefits and risks of recreational running, we sent questionnaires to 1,250 randomly selected male and 1,250 female registrants for a 10-km road race. The response rate was 55% for men and 58% for women. Telephone interviews of a randomly selected group of nonrespondents indicated that the only significant differences between respondents and nonrespondents were that (1) respondents were older than nonrespondents, (2) more male nonrespondents had stopped running during the year after the race, and (3) more male nonrespondents had been hit by thrown objects. One year after the race, 89% of male and 79% of female respondents were still running regularly. Eighty-one percent of men and 75% of women who smoked cigarettes when they began running had stopped smoking after beginning recreational running. Giving up smoking was significantly more common for current runners than for "retired" runners. Weight loss was commonly associated with running and was greater in those persons who were overweight when they began running. More than a third of respondents had a musculoskeletal injury attributed to running in the year after the race and about one seventh of all respondents sought medical consultation for their injury. The risk of injury increased with increasing weekly mileage. This study uses epidemiologic methods to quantify some of the benefits and risks of running.