For a correct interpretation of analytical results in doping control, knowledge on the stability of prohibited substances in the urinary matrix is a prerequisite. So far, limited data is available on the stability of prohibited substances in unaltered urine because most of the studies investigating the stability of drugs have used stabilized, sterilized, or filtered urine. In this work, the long-term stability of ephedrine, methylephedrine, cathine, 19-norandrosterone glucuronide, and a wide range of diuretics was determined over a period of 9 months at -20 degrees C, 4 degrees C, 22 degrees C, and 37 degrees C. Short-term stability, including the influence of 6 freeze-thaw cycles and 15 h storage at 60 degrees C, was also investigated. Often, a tolerance limit of 15%, similar to what is commonly used in the evaluation of precision data during method validation, is used to evaluate stability. This paper describes an alternative approach, using measurement uncertainty data to evaluate long-term stability with a probability of 95%, and proposes a simple alternative for investigating the stability for non-threshold substances. The results indicate that all the investigated substances are stable (alpha=0.05) when stored at -20 degrees C and 4 degrees C, but that at higher temperatures significant degradation effects can occur. The study also shows that degradation can be dependent on the urinary matrix and that the results from stability studies using stabilized, filtered, or sterilized urine can underestimate degradation effects.