An in vitro P-glycoprotein mediated drug biliary excretion model (B-Clear model) was developed and validated using sandwich-cultured rat hepatocytes (SCRH) and a model substrate rhodamine 123 (Rh123).  SCRH formed functional bile canalicular networks after 5 days of culture.  Rh123 (10 μmol·L⁻¹) was then incubated with the SCRH in standard Ca²⁺ Hanks buffer or Ca²⁺-free buffer.  The cumulative cell uptake and canalicular efflux of Rh123 under Ca²⁺ and Ca²⁺-free conditions were measured with a LC-MS/MS method.  The biliary excretion index (BEI) and instinct biliary clearance (CL_{bile, int}) were calculated.  To assess the effect of known P-gp inhibitors on the efflux of Rh123, cyclosporine A (CyA), tariquidar (TQD) or quinidine (QND) (10, 50 and 100 μmol·L⁻¹) was pre-incubated separately with SCRH for 30 min, then co-incubated with Rh123.  The BEI and CL_{bile, int} of Rh123 obtained from the SCRH model were (17.8 ± 1.3) % and (10.7 ± 0.9) mL·min⁻¹·kg⁻¹, respectively.  All the three P-gp inhibitors showed a dose-dependent inhibition on the bile clearance of Rh123, indicating that the B-Clear model with SCRH was functional properly.  The biliary excretion of loperamide (LPAD) and the role of P-gp were further investigated with this validated model.  The BEI and CL_{bile, int} for LPAD (20 μmol·L⁻¹) were obtained after it was incubated with SCRH for 30 min, and found to be (12.9 ± 1.2) % and (6.1 ± 0.3) mL·min⁻¹·kg⁻¹ respectively.  The dose-dependent inhibition on LPAD biliary excretion by CyA, TQD or QND confirmed the major role of P-gp in LPAD canalicular efflux.  The results suggested that the B-Clear model with SCRH would be a useful tool for evaluation of P-gp mediated efflux and drug-drug interaction.