Studies on the use of performance enhancing drugs (PED) in fitness centres rely predominately on conventional survey methods using direct questioning. However, research indicates that direct questioning of sensitive information is characterized by under-reporting. The aim of the present study was to contrast direct questioning of different types of PED use by Dutch fitness centre members with results obtained with the Randomized Response Technique (RRT). Questionnaires were conducted among members of fitness centres. PED were classified into the following categories: anabolic steroids, prohormones, substances to counteract side-effects, growth hormone and/or insulin, stimulants (to reduce weight), and miscellaneous substances. A total of 718 athletes from 92 fitness centres completed the questionnaire. The conventional method resulted in prevalences varying between 0% and 0.4% for the different types of PED with an overall prevalence of 0.4%. RRT resulted in prevalences varying between 0.8% and 4.8% for the different types of PED with an overall prevalence of 8.2%. The overall prevalence of the two survey methods differed significantly. The current study showed that the conventional survey method using direct questioning led to an underestimation of the prevalence. Based on the RRT results, the percentage of users of PED among members of fitness centres is approximately 8.2%. Stimulants to lose weight had the highest prevalence, even higher than anabolic steroids. The key task for future preventive health work is to not only focus on anabolic steroid use, but also include interventions focusing on the use of stimulants to lose weight. © 2013 John Wiley & Sons, Ltd.