The methods in use in the authors laboratory for the determination of plasma testosterone, androstenedione and dihydrotestosterone are described and possible pitfalls are discussed. The gas-chromatographic methods are compared to methods based on competitive protein binding and the pros and cons of each are discussed. It appears that in expert hands GLC methods for plasma androgens are not inferior to methods based on CPB, the only advantage of the latter being their greater practicability, which is however largely neutralized by the lack of specificity of CPB methods, requiring extensive preparatory purification of material used.