PURPOSE: To objectively compare side-to-side patellar position and mobility in patients with idiopathic unilateral anterior knee pain (AKP) using a stress-testing device concomitantly with magnetic resonance imaging. It is hypothesized that the painful knees present greater patellar mobility than the contralateral non-painful knees. METHODS: From a total sample of 359 patients, 23 patients with idiopathic unilateral AKP (30.9 years, 23.4 kg/m(2), 43% males) were included within the present study. Both knees of all the patients were examined by conventional imaging, including the measurement of trochlear sulcus angle, Caton-Deschamps index, tibial tuberosity to trochlear groove distance, patellar tilt angle and patellar subluxation (both at rest and upon quadriceps contraction). Additionally, the same patients underwent stress testing (Porto Patella Testing Device); these measurements were taken with the patella at rest, after lateral patellar translation and after lateral patellar tilt. Clinical and functional outcomes were obtained using physical examination and the Kujala and Lysholm scores. RESULTS: Painful knees showed statistically significant higher patellar lateral position after stressed lateral translation than non-painful knees (p = 0.028), 9.8 ± 3.6 mm and 7.1 ± 6.3 mm, respectively. The adjusted multivariate logistic model identified the patellar position after lateral displacement to be significantly associated with AKP (OR = 1.165) and the model (AUC = 0.807, p < 0.001) showed reasonable sensitivity (67%) and specificity (73%). CONCLUSION: Patients with idiopathic unilateral AKP with morphologically equivalent knees showed statistically significant increased patellar lateral position after stressed lateral displacement in their painful knee. The greater lateral patellar mobility quantified by the PPTD testing brings more objectivity to the diagnosis. LEVEL OF EVIDENCE: II.