Objectives: To evaluate the repeatability and reproducibility of ultrasonographic femoral trochlear depth measurements and to compare ultrasonographic and intra-operative evaluations of femoral trochlear depth for predictive value in selecting trochleoplasty. Methods: Repeatability and reproducibility of an ultrasonographic protocol were tested in a preclinical cadaveric study. Clinical patients undergoing corrective surgery for patellar luxation were evaluated preoperatively with ultrasound and intra-operatively using a depth gauge. Measurements were assessed for equivalence using linear regression, and agreement between decisions made based on these measurements assessed using Cohen’s kappa. Results: Although ultrasonographic and intra-operative measurements were in broad agreement, the prediction interval was too wide for clinical use. There was no significant agreement between predictions of the need for trochleoplasty using various cut-off values for the two measurements, nor between these and the surgeon’s decision. Clinical significance: Based on our observations, use of ultrasound for evaluation of the femoral trochlea remains a largely qualitative assessment. A simpler and more direct objective measure of femoral trochlear adequacy is required for intra-operative use.