Aim To examine the effects of a rural community-based program integrating health education and self-directed home-based resistance training on the mobility function of older women with knee osteoarthritis (OA). Methods Participants were assigned to the control (CON) or intervention (INT) group. Participants completed a mobility function assessment and questionnaire evaluation. The CON group received only the health education program and the INT group also participated in resistance training for 5 months. Primary outcomes were mobility function measured using a timed chair stand (TCS), timed up & go (TUG), gait speed (GS), and knee extensor strength (KES). We evaluated body composition and questionnaire results (WOMAC score, SARC-F scale, and SOF index) as secondary outcomes. Results There were significant differences in the interactions of group-by-time effects for TCS ( p  < 0.001), TUG ( p  = 0.006), GS ( p  = 0.020), and knee strength ( p  = 0.010). In the CON group, TCS ( p  = 0.003) and TUG ( p  = 0.005) increased compared with baseline, while in the INT group, TCS decreased significantly ( p  < 0.001) and TUG tended to decrease after the intervention. The INT group showed improvement in GS ( p  < 0.001) and KES ( p  = 0.003) after the intervention. By contrast, the CON group showed decreasing GS ( p  = 0.021) and KES ( p  = 0.011) compared with baseline. The SARC-F scale differed significantly in the interaction of group-by-time effects ( p  = 0.030), while the body composition, SOF index, and WOMAC score did not differ. Conclusion These results suggest that an integrated intervention program combining self-directed home-based resistance training with health education effectively improves the mobility function of older adults with knee OA dwelling in rural areas.