Population pharmacokinetic/pharmacodynamic (PK/PD) models were developed to quantitate the exposure–response relationships using continuous longitudinal data on American College of Rheumatology (ACR) subcomponents, that is, tender-joint count (TJC), swollen-joint count (SJC), C-reactive protein, patient's assessment of pain, patient's global assessment of disease activity, physician's global assessment of disease activity, and patient's assessment of physical function for 5 biologics approved for use in rheumatoid arthritis. The models were then used to simulate the time courses of clinical outcomes following different treatment regimens. The relative sensitivity of the 7 subcomponents was assessed using Monte Carlo simulation–based power analysis. The developed population PK/PD models adequately described the relationship between serum concentrations and changes in ACR subcomponents. The trial simulation and subsequent power analysis showed that SJC and TJC appeared to be more sensitive than the other 5 ACR subcomponents to detect treatment effect over placebo/methotrexate. These 7 ACR subcomponents had similar power in detecting the treatment difference between different doses. In addition, the continuous measures of ACR subcomponents did not appear to be more sensitive than binary measures.