Background <p>This study investigated the relationship between the vastus medialis obliquus (VMO) cross-sectional area (CSA) and patellofemoral instability (PFI) in both primary and recurrent lateral patellar dislocations (LPD). Our secondary objective was to examine associations between VMO CSA and trochlear dysplasia, tibial tuberosity position, and patellar height in patients with PFI.</p> Methods <p>Magnetic resonance imaging (MRI) radiographs were retrospectively analysed for 90 patients with primary acute LPD, 90 patients with recurrent LPD, and 56 patients without LPD (control). Measurements of the CSA ratio of the VMO to the whole thigh in three transverse slices were performed to calculate a mean ratio per patient. Additionally, tibial tubercle-trochlear groove (TT-TG) distance, patellar tilt angle (PTA), trochlear sulcus angle (TSA), and Insall-Salvati ratio (ISR) were measured as part of the Dejour Protocol.</p> Results <p>The median CSA ratios in primary (0.04, standard deviation [SD]: 0.02) and recurrent (0.04, SD: 0.02) LPD patients were significantly lower than those in the control group (0.07, SD: 0.02) (<i>P</i> &lt; 0.05). Compared with the primary LPD group, the recurrent LPD group presented significantly greater TT-TG distances (16.0, SD: 4.77&#xa0;mm vs. 13.0, SD: 4.73&#xa0;mm; <i>p</i> = 0.0101) and PTA (25, SD: 9.79 degrees vs. 19, SD: 15.76 degrees; <i>p</i> = 0.0071). There was no statistically significant correlation between any parameters of the Dejour Protocol and the VMO CSA ratio in patients with primary or recurrent dislocations (<i>P</i> &gt; 0.05).</p> Conclusion <p>Patients with both primary and recurrent LPD demonstrated smaller VMO bulk relative to the rest of the thigh compared with controls. These findings indicate an association between reduced VMO size and patellar dislocation; however, causality cannot be inferred from this cross-sectional analysis.</p> Level of evidence <p>IV.</p>