ONO-5334, a selective inhibitor of cathepsin K, is a potential new treatment for osteoporosis. The objectives of this modeling study were to (1) develop exposure–response (E–R) models to relate ONO-5334 exposure to bone mineral density (BMD), (2) predict BMD responses to various doses of ONO-5334 for both immediate release tablet (IRT) and sustained release tablet (SRT) formulations where only BMD response after administration of IRT had been studied to date, (3) inform selection of appropriate formulation/dose using simulation for future clinical trials. A population pharmacokinetic (PK) model was developed to simultaneously analyze data for both IRT and SRT. The exposure metrics at steady state were estimated by post hoc Bayesian prediction using the final population PK model. E–R models were developed using dose-ranging data with only IRT from postmenopausal females with osteoporosis. Based on the developed model, lumbar spine and total hip BMD after administration of ONO-5334 SRT as well as IRT were simulated. The simulation results showed that ONO-5334 SRT should provide comparable BMD responses at a lower dose relative to IRT (a finding consistent with the results from a previous population PK–PD modeling study with bone resorption markers).