Two clinical studies were carried out: a smoking study and a controlled oral administration study. Hemp was ingested either as a water or a milk decoction. Because of its lipophylic character, much higher levels of THC were extracted into milk compared to water during the tea-making process. The highest mean concentrations of THC, 11-OH-THC and THCCOOH were 4.0, 3.4 and 24.5 ng/ml whole blood after drinking the milk decoction. Trace amounts of THCCOOH only could be measured in the blood after ingestion of the water tea and this without any perceptible effects typical of cannabis use. Unlike the water tea, ingestion of 23.2 mg of THC with the milk decoction resulted in significant psychoactive and clinical effects. The time since exposure was estimated with the mathematical models I and II set up by Huestis et al. and the results compared to the actual time of use. In our set of experimental conditions, model I yielded much better predictions of time of cannabis exposure than model II. Insofar as an oral intake is suspected, e.g. when the THC/11-OH-THC ratio is close to one, it is recommended to use model I exclusively. Alternatively, model II modified to take account of our own set of data could be also used to achieve an acceptable time estimate.