Only 10% of all compounds developed by pharmaceutical companies make it to the market. Of the 90% that do not make it to the market, 50% either have toxicity or pharmacokinetic issues. Thus, the need for ADMET (absorption, distribution, metabolism, excretion and toxicity) optimization during the early stages of drug development is clear. In silico tools may be promising for this use due to their lower cost and time requirements. This review aims to evaluate the predictive power of intestinal absorption and oral bioavailability prediction methods using different statistical approaches over time. Improvement, refinement and diversification of these methods have been observed over the past few years. Nevertheless, some elements related to the quality of the biological data, disclosure of the data used and description of validation methods, that could contribute to building new, better and more reliable models have been ignored by researchers or restricted by the technical limitations of various laboratories.