Purpose <p>To assess the utility of pre-procedural knee MRI prior to genicular artery embolization (GAE) for the prognostication of early outcomes for symptomatic knee osteoarthritis (KOA).</p> Materials and Methods <p>A single-center study including 39 patients received GAE from 9/2021–4/2025 with pre-procedural MRIs. MRIs were evaluated for structural abnormalities including menisci, ligaments, cartilage, marrow signal, and loose bodies. For 24 patients, synovitis was assessed using a semiquantitative method. Clinical outcomes were measured with the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) scoring system at pre-GAE and three-month postintervention. Categorical response was assessed as a 50% reduction in the WOMAC Pain. Student t-tests were used to evaluate WOMAC pain reduction, and subset analysis was performed between MRI structural abnormalities and categorical response.</p> Results <p>Fifty-two knees received GAE with pre-procedural MRIs (33 with contrast). There was a 34.6% clinical response rate (<i>N</i> = 8/52). Lateral meniscus and cartilage abnormality were predictive of poor categorical response at three months (<i>P</i> = 0.039–0.040). ≥ 4 structural abnormalities were associated with poor treatment response (<i>P</i> = 0.004). Pre-GAE synovitis was not predictive of categorical response at three months (<i>P</i> = 0.809). Kellgren–Lawrence ≥ 3 was predictive of poor response (<i>P</i> &lt; 0.001). Lower adverse event rate was observed with temporary embolic compared to permanent embolic (<i>P</i> = 0.032).</p> Conclusion <p>Pre-GAE knee MRI may offer short-term prognostic utility. Pre-procedural abnormalities in the lateral menisci and cartilage can predict poor response to GAE at three months. A greater degree of structural abnormality (≥ 4 structural abnormalities) was associated with poor response. Temporary embolic agent may be safer than permanent embolic agent.</p> Graphic Abstract <p></p>