A bioanalytical method for the analysis of artesunate (ARS) and its metabolite dihydroartemisinin (DHA) in human plasma using protein precipitation and liquid chromatography coupled to positive tandem mass spectroscopy was developed. The method was validated according to published US FDA-guidelines and showed excellent performance. However, when it was applied to clinical pharmacokinetic studies in malaria, variable degradation of the artemisinins introduced an unacceptable large source of error, rendering the assay useless. Haemolytic products related to sample collection and malaria infection degraded the compounds. Addition of organic solvents during sample processing and even low volume addition of the internal standard in an organic solvent caused degradation. A solid phase extraction method avoiding organic solvents eliminated problems arising from haemolysis induced degradation. Plasma esterases mediated only approximately 20% of ex vivo hydrolysis of ARS into DHA. There are multiple sources of major preventable error in measuring ARS and DHA in plasma samples from clinical trials. These various pitfalls have undoubtedly contributed to the large inter-subject variation in plasma concentration profiles and derived pharmacokinetic parameters for these important antimalarial drugs.