There are no data on the efficacy of a back school in primary prevention of back pain in the general population or on the characteristics of the population who volunteers. After announcement in the local press, 494 healthy adults volunteered and paid for a back school course in Switzerland. A total of 371 controls were matched for sex, age, profession, nationality and back pain. A statistically significant decrease in numbers of doctor's visits was found by the participants during the following 6 months compared with the controls. However, there were no significant between-group difference in the four remaining parameters (presence and intensity of back pain, drug intake and sick leave). Three-quarters of participants changed their attitudes after the back school. Volunteering for a back pain prevention programme was associated with the presence of back pain problems. Reasons for volunteering are further discussed. Overall, the results of this study showed that a back school for the general population may not solve the problem of low back pain, but improves self-help in a subgroup of the population.