OBJECTIVE To describe the relationship between lower extremity physical performance, self-reported mobility difficulty, and self-reported use of compensatory strategies (CSs) for mobility inside the home. DESIGN Cross-sectional exploratory study. SETTING Community-dwelling elders. PARTICIPANTS Disabled, cognitively intact women 65 years or older (N=1002), from the Women's Health and Aging Study I. INTERVENTIONS Not applicable. MAIN OUTCOME MEASURES CS scale: no CS, behavioral modifications (BMs) only, durable medical equipment (DME) with or without use of BMs, and any use of human help (HH); and 3 dichotomous CS measures: any CS (vs none); DME+HH (vs BMs only, among users of any CS); any HH (vs DME only, among users of any DME/HH). RESULTS Self-reported mobility difficulty and physical performance were significantly correlated with one another (r=-.57, P<.0001) and with the CS scale ([r=.51, P<.001] and [r=-.54, P<.0001], respectively). Sequential logistic regressions showed self-reported difficulty and physical performance were significant independent predictors of each category of CS. For the any CS and DME+HH models, the odds ratio for self-reported difficulty decreased by approximately 50% when physical performance was included in the model, compared with difficulty alone ([18.0 to 8.6] and [7.3 to 3.8], respectively), but both physical performance and difficulty remained significant predictors (P<.0001). The effects of covariates differed for the various CS categories, with some covariates having independent relationships to CS, and others appearing to have moderating or mediating effects on the relationship of self-reported difficulty or physical performance to CS. CONCLUSIONS Physical performance, self-reported difficulty, health conditions, and contextual factors have complex effects on the way elders carry out mobility inside the home.