Purpose: Knee osteoarthritis (OA) is one of the most common causes of disability in the aging population. Prevention of progression of the disease is a key objective. Hence, knowledge of factors associated with progression of knee osteoarthritis is of paramount importance as only a limited number of the patients progress rapidly to end-stage OA with total knee arthroplasty. Few longitudinal studies have investigated prognostic factors for both structural, clinical and functional deterioration in patients with medial knee OA. In this 2 year follow up study, baseline characteristics of a group of progressors, based on distinct structural, clinical, and functional outcomes, was compared to non-progressors. Methods: In this longitudinal study, we investigated a cohort of women with both early and established medial knee OA (n = 57) over two years. We classified subjects into progressors and non- progressors, first, based on their structural changes as detected on plain radiography (any worsening in the grade for radiographic medial joint space K&L grading); second, based on their clinical change in pain (10 points on the KOOS pain subscale); third, based on their changes in function as assessed by self-reported physical function of KOOS (a 10 points drop); and lastly, based on their changes in objective performance-based physical function a 20% decline (more time spent to perform the task after 2 years) in the Timed up and Go test (TUG) and stair climbing test (SCT). We compared the baseline characteristics between progressors and non-progressors, for each category, separately. Differences in MRI features between the fast and slow progressors were tested with Fisher exact test or Mann-Whitney U test. To test for group differences between fast and slow progressors for baseline measures of static alignment as well as for the KOOS subscores, the Mann-Whitney U test was used. For BMI, dynamic alignment, muscle strength, and dynamic loading an independent sample’s t-test was used. Results: Seventeen subjects (30%) with either early or established medial knee OA worsened in the K&L score after two years compared to baseline, and were categorized as structural progressors. These patients showed at baseline more knee pain and symptoms, as well as, worse self-reported Activity Daily Living and Quality of Life as measured with KOOS. On MRI, they displayed significantly more and larger tibiofemoral BMLs, more meniscal extrusion and meniscal maceration. In addition, a pattern towards a higher second peak knee adduction moment during stance phase of gait, at baseline, was found in the progressors’ group compared to the non-progressors (P = 0.071). After 2 years, eleven patients (23%) showed increased self-reported pain levels as measured by KOOS pain subscale and ten patients (21%) showed progression based on decline in self-reported physical function (ADL). Those progressors (based on pain or self-reported function) had lower isometric and isokinetic quadriceps strength and isometric hamstring strength as compared to non-progressors. As for objective performance, eight patients showed a 20% decline in TUG/SCT after 2 years. The percentage of subjects with meniscal tears was higher in the progressors group compared to non-progressors group. No other structural, clinical, or functional characteristics were significantly different between groups at baseline. Conclusions: In this longitudinal cohort study of women with symptomatic medial knee OA, we found that 20–30% of subjects did progress over 2 years based on either structural worsening, pain exacerbation, or functional deterioration. Our data indicate that the baseline patient characteristics associated with the distinct outcomes appear not to be the same. Our findings may contribute to ongoing efforts to identify “a patient at risk” at the time of diagnosis with early or established OA.