The method described allows quantitative analysis of endogenous level down to 0.2 ng harman and 2 ng norharman in 1 mL plasma or 10 to 8 thrombocytes. The levels of harman detected and quantified in smokers and non-smokers, b ut also in a patient treated with chloral do not excedd the blanks. These considerations together with the results of this investigation lead to the conclusion that albeit beta-carbonlines like harman and norharman exert manifold effects on neurotransmitter systems of the brain but also on peripheral enzym complexes, it is doubtful whether they occur in the human organism in amounts of physiological sign ificance.