To quantify the drug–drug interactions between dabigatran etexilate (DE) and proton pump inhibitors (PPI) and in particular the role of P-gp activity modulation. In the first part of the study, efflux ratios of DE were evaluated using the caco-2 cell line in the presence of pantoprazole, omeprazole, rabeprazole, lansoprazole and ciclosporin A (positive control). The two PPI that reduced the efflux ratio of dabigatran to the greatest and least extent, respectively, were used during the second part of the study, comprising a single-centre, randomised, open-label study with an incomplete Latin square design. Nine healthy volunteers received DE (150 mg) alone, DE (150 mg) with the first PPI and DE (150 mg) with the second PPI in randomised sequence. Dabigatran plasma concentration and thrombin time were measured in blood samples withdrawn at 11 time points after each treatment. Models were built using a nonlinear mixed-effect modelling approach. Omeprazole and rabeprazole were the two PPI that reduced the efflux ratio of DE least and most, respectively. The PK model was based on an inverse Gaussian absorption process with one compartment. The relationship between dabigatran concentration and thrombin time was considered linear. Some PK profiles had dramatically low concentration values due to poor absorption. These profiles were clustered using a between subject model mixture with interoccasion variability. The concomitant administration of PPI did not significantly change dabigatran pharmacokinetics. DE is subject to high absorption variability, precluding evaluation of the effect of PPI on its pharmacokinetics.