Background: Patients treated with sunitinib show high inter-patient variability in drug exposure (40-60%), which is largely unexplained. Since sunitinib is metabolized by CYP3A4, variability in the activity of this enzyme may explain a considerable proportion of the observed variability. We therefore prospectively studied the relationship between CYP3A4 activity and systemic exposure to sunitinib. Methods: In fifteen patients treated with sunitinib in a four weeks “on” – two weeks “off” regimen the pharmacokinetics of sunitinib and its active metabolite SU12662 were assessed. To determine sunitinib+SU12662 steady-state exposure, samples were collected over 24hrs after at least 14 days of sunitinib therapy. To assess CYP3A4 activity, midazolam 7.5mg orally was administered on the final day of the two weeks “off”. Plasma concentrations were measured over a period of 7hrs to determine midazolam exposure. Exposures (AUC) were calculated using a trapezoidal approach (Phoenix WinNonlin v6.3). The relationship between CYP3A4 activity (midazolam exposure) and sunitinib+SU12662 exposure was determined by linear regression analysis. The percentage of variability in sunitinib+SU12662 exposure that could be explained by CYP3A4 activity was calculated by Pearson’s correlation. In addition, the correlation between sunitinib+SU12662 C_trough levels and sunitinib+SU12662 exposure was assessed. Results: A strong correlation between midazolam exposure (AUC_0-7hr) and steady-state sunitinib+SU12662 exposure (AUC_0-24hr) was found (p= 0.002); CYP3A4 activity explained 55% of the observed inter-patient PK variability of sunitinib+SU12662. Furthermore sunitinib+SU12662 C_trough levels were highly predictive (96%) for overall sunitinib+SU12662 exposure (AUC_0-24hr). Conclusions: Midazolam as a phenotyping probe could be useful before start of sunitinib therapy to identify patients at risk for under- respectively overtreatment at a standard dosage regimen. Therefore, CYP3A4 phenotyping could be useful to individualize sunitinib therapy. Additionally, sunitinib+SU12662 trough levels are highly predictive for sunitinib+SU12662 exposure and thus can be used for monitoring and guiding sunitinib therapy in clinical practice. Clinical trial information: NCT01743300.