ABSTRACT Purpose In competitive sport, classic methods of measuring drug prevalence, such as doping controls or questionnaires, are challenging. Here we describe a novel urine sampling method to measure drug use in athletes. We hypothesize that the prevalence of drug use in ultramarathon runners is measured more accurately with our sampling method than randomized-response questionnaires. Methods Urine samples and associated demographic data were collected from male participants using blind, automated urinals at the start of ultramarathon races. Various non-prohibited and prohibited substances were subsequently screened. Concomitantly, 2,931 male and female runners participating in the same ultramarathons completed an anonymized, randomized-response questionnaire regarding drug use. Results Among 412 individual urine samples, 205 (49.8%) contained at least one substance, and 16.3% of the samples contained one or more prohibited substances. Substances detected in urine included non-steroid anti-inflammatory drugs (NSAIDs) (22.1%), acetaminophen (15.5%), opioids (6.6%), diuretics (4.9%), hypnotics (4.4%), glucocorticoids (2.7%), beta-2 agonists (2.2%), cannabinoids (1.9%) and stimulants (1.2%). None of the samples contained erythropoietin-receptor agonists or suspicious testosterone. Drug use was not associated with the participants’ characteristics or ranking. Respondents to the questionnaire reported using acetaminophen (13.6%) and NSAIDs (12.9%); however, no prohibited substances were declared. Conclusions There was a high prevalence of drug use among male ultramarathon runners, in particular NSAIDs and painkillers; however, performance-enhancing drugs were marginally used. Blind urine sampling highlighted prohibited drug use not declared in questionnaires, and it is useful to assess the prevalence of drug use and/or doping in competitive athletes.