Bioavailability data for herbal supplements in humans is not readily available or is difficult to obtain, because of the complexity of the composition and the diversity of the constituents. Potency of an herbal extract is due to the synergistic interactions between several constituents. Thus, the use of in silico methods is an attractive alternative to predict the qualitative intestinal permeability of the active constituents for the selection of appropriate bioavailability markers. Molecular descriptors such as CLogP, minimal cross-sectional area and polar surface area of 37 active components from selected herbal extracts such as milk thistle, kava, ginkgo, ginseng, valerian, black cohosh and garlic were estimated. In vitro permeability of the compounds was determined by SimBioDAS an in vitro epithelial cell permeability assay. Based on the in silico descriptors and their relationship with the in vitro permeability, the qualitative intestinal permeability of the active compounds was predicted. Bioavailability and bioequivalence markers were predicted for kava, Ginkgo biloba and milk thistle. Choosing a compound which has the least intestinal permeability as a marker is the most conservative approach toward ensuring the bioavailability of the entire extract.