OBJECTIVES: To explore the dimensionality of the PF construct when HAQ-DI and PF10 items are combined into a single instrument. METHODS: Item Response Theory (IRT) was used to analyze data from a phase 3, randomized, placebo-controlled study in patients with RA. Dimensionality of the combined instrument was explored by factor analyses (exploratory and confirmatory).  Two approaches were taken to address the multi-factor/dimensionality issue identified through factor analysis. The first approach was to fit a sequence of unidimensional Rasch and generalized partial credit (GPC) models to each factor. Model comparisons and diagnoses were performed to assess the adequacy of model fitting and to detect any violation of the underlying assumptions. The second approach was to fit a multi-dimensional IRT model to these items.  RESULTS: The factor analysis suggested there were three underlying factors in the combined PF10 and HAQ-DI. These three factors can be best characterized as hygiene (including 4 items from HAQ-DI and 1 from PF10), lower body activity (including 8 items from HAQ-DI and 6 from PF10) and upper body activity (including 7 items from HAQ-DI and 3 from PF10).  Overall, the GPC model significantly outperformed the Rasch model across all factors (all p-value<0.0001), implying divergent discriminability among items within each factor. The multi-dimensional IRT improved the model fit but substantially increased the computation effort and sacrificed model interpretability. CONCLUSIONS: The pooled HAQ-DI and PF10 items represent multiple domains of physical functioning. To explain the combination of these items, a multi-dimensional IRT model decreased model interpretability and increased computation intensity compared to unidimensional models.