1. Homozygously mdr1a gene disrupted mice (mdr1a(-/-) mice) and wild type mice (mdr1a(+/+) mice) were used to develop a method for P-glycoprotein (P-gp) function imaging non-invasively and to study the effect of a P-gp reversal agent on its function in vivo. 2. [11C]verapamil (0.1 mg/kg) was administered and the changes in tissue concentrations were determined ex vivo by organ extirpation and in vivo with PET. To block P-gp function, cyclosporin A was administered. 3. Biodistribution studies revealed 9.5-fold (P < 0.001) and 3.4-fold (P < 0.001) higher [11C]verapamil in the brain and testes of mdr1a(-/-) mice than in mdr1a(+/+) mice. Cyclosporin A (25 mg/kg) increased [11C]verapamil levels in the brain and testes of mdr1a(+/+) mice in both cases 3.3-fold (P < 0.01 (brain); P < 0.001 (testes)). Fifty mg/kg cyclosporin A increased [11C]verapamil in the brain 10.6-fold (P < 0.01) and in the testes 4.1-fold (P < 0.001). No increases were found in the mdr1a(-/-) mice. This indicates complete inhibition of P-gp mediated [11C]verapamil efflux. 4. Positron camera data showed lower [11C]verapamil levels in the brain of mdr1a(+/+) mice compared to those in mdr1a(-/-) mice. [11C]verapamil accumulation in the brain of mdr1a(+/+) mice was increased by cyclosporin A to levels comparable with those in mdr1a(-/-) mice, indicating that reversal of P-gp mediated efflux can be monitored by PET. 5. We conclude that cyclosporin A can fully block the P-gp function in the blood brain barrier and the testes and that PET enables the in vivo measurement of P-gp function and reversal of its function non-invasively.