Antiresorptive medications have been explored for treating knee osteoarthritis (OA);​ however, little data exist on the effects of today's more potent nitrogen‐containing oral bisphosphonates on radiographic disease‐progression in patients with varying disease‐severity, especially those who are not overweight. The primary objective of this cohort study was to determine if the use of bisphosphonates is protective against 2‐year radiographic‐progression of knee OA in Osteoarthritis Initiative (OAI) participants, stratified by baseline radiographic disease status. Secondary objectives were to examine effects in non‐overweight participants (body mass index [BMI] < 25 kg/m²) and cumulative bisphosphonate exposure effects. We identified female OAI participants aged ≥50 years and excluded those missing baseline radiograph readings, bisphosphonate use information, or all clinical questionnaire information at baseline. Participants reporting bisphosphonate use (69% alendronate) were propensity‐matched 1:​1 to non–bisphosphonate users and followed until first radiographic knee OA progression (1‐unit increase in Kellgren and Lawrence [KL] grade) or data were censored (first missed visit or end of 2‐year follow‐up). Discrete‐time logistic regression models estimated hazard ratios (HRs) between bisphosphonate users versus nonusers, with an interaction term for baseline KL grade (KL <2 or KL ≥2). We identified 1977 eligible women (n = 346 bisphosphonate users). Propensity‐matched results indicated that bisphosphonate users with KL grade <2 were protected against progression (HR_KL<2 0.53;​ 95% CI, 0.35 to 0.79), while bisphosphonate use was not associated with radiographic progression in those with KL grade ≥2 (HR_KL≥2 1.06;​ 95% CI, 0.83 to 1.35). When restricting analyses to those with BMI <25 kg/m², effects were strengthened (HR_KL<2 0.49 [95% CI, 0.26 to 0.92];​ HR_KL≥2 0.69 [95% CI, 0.33 to 1.26]). Duration of bisphosphonate use had no effect on progression, though sample size was limited. Bisphosphonate therapy may be protective against radiographic knee OA progression in early‐stage patients, particularly those who are non‐overweight, but less so for those with more advanced disease or more weight‐bearing joint stress.