The population pharmacokinetic approach has proved its usefulness for Therapeutic Drug Monitoring (TDM) and it allows in particular, the quantification of pharmacokinetic variability, and offers the possibility to detect clinically useful predictors of treatment outcomes. In that perspective, we performed a first pharmacogenetic-based population pharmacokinetic analysis of efavirenz in HIV infected individuals. We demonstrated that functional alleles of CYP2B6 were associated with higher Efavirenz exposure. We proposed an innovative non-linear model to describe the relationship between phenotypes and functional allelic variants and accordingly predict Efavirenz exposure. We developed Pop-PK models for darunavir, etravirine and raltegravir. We identified demographic and environmental factors affecting drug disposition. We simulated compare drug exposure under several dosage regimens. Finally, we derived reference pharmacokinetic curves for 7 widely used antiretroviral drugs based on an innovative concept of meta-analysis that could represent useful tools for clinicians to interpret the value of an individual plasma concentration whenever needed.