Background Popliteal sciatic nerve catheters (PSNCs) are associated with a high frequency of displacement. We aimed to estimate the frequency of catheter displacement after 48 h with magnetic resonance imaging (MRI) in patients with PSNCs after major foot and ankle surgery randomized to catheter insertion either with a short-axis in-plane (SAX-IP) approach perpendicular to the nerve or with a short-axis out-of-plane (SAX-OOP) approach parallel to the nerve. Methods Forty patients were randomly allocated to SAX-IP or SAX-OOP PSNC. Ropivacaine 0.75% 20 ml was injected via the catheter followed by ropivacaine 0.2% 10 ml h-1 infusion. Correct primary catheter placement was ensured after initial injection of local anaesthetic via the catheter. Forty-eight hours after insertion, MRI was performed after injection of saline with added contrast (Dotarem) via the catheter. The primary outcome was catheter displacement estimated as the frequency of spread of contrast exclusively outside the paraneurium. Results All patients had correct primary catheter placement. The frequency of displacement 48 h after insertion of the PSNC was 40% when inserted perpendicular to the nerve vs 10% parallel to the nerve (difference was 30 percentage points, 95% CI: 3-53 percentage points). The relative risk of displacement was four times larger (95% CI: 0.8-10, P<0.028) in the SAX-IP vs the SAX-OOP group. The morphine consumption was 150% greater in the SAX-IP compared with the SAX-OOP group. Conclusion Popliteal sciatic nerve catheters for major foot and ankle surgery inserted with ultrasound guidance parallel to the sciatic nerve have a significantly lower frequency of displacement compared with those inserted perpendicular to the nerve. Clinical trial registration NCT02200016.