PURPOSE: Multiple-ligament knee reconstruction techniques often involve the creation of several bone tunnels for various reconstruction grafts. A critical step in this procedure is to avoid short tunnels or convergences among them. Currently, no specific template guide to reproduce these angulations has been reported in the literature, and the success of the technique still depends on the experience of the surgeon. The aim of this study is to analyze the accuracy and reliability of 3D-printed patient-specific instrumentation (PSI) for lateral and medial anatomical knee reconstructions. METHODS: Ten cadaveric knees were scanned by computed tomography (CT). Using specific computer software, anatomical femoral attachments were identified: (1) on the lateral side the lateral collateral ligament (LCL) and the popliteal tendon (PT) and (2) on the medial side the medial collateral ligament (MCL) and the posterior oblique ligament (POL). Four bone tunnels were planned for each knee, and PSI with different directions were designed as templates to reproduce the planned tunnels during surgery. Twenty 3D-printed PSI were used: ten were tailored to the medial side for reconstructing MCL and POL tunnels, and the other ten were tailored to the lateral side for reconstructing LCL and PT tunnels. Postoperative CT scans were made for each cadaveric knee. The accuracy of the use of 3D-printed PSI was assessed by superimposing post-operative CT images onto pre-operative images and analyzing the deviation of tunnels performed based on the planning, specifically the entry point and the angular deviations. RESULTS: The median entry point deviations for the tunnels were as follows: LCL tunnel, 1.88 mm (interquartile range (IQR) 2.2 mm); PT tunnel, 2.93 mm (IQR 1.17 mm); MCL tunnel, 1.93 mm (IQR 4.26 mm); and POL tunnel, 2.16 mm (IQR 2.39). The median angular deviations for the tunnels were as follows: LCL tunnel, 2.42° (IQR 6.49°); PT tunnel, 4.15° (IQR 6.68); MCL tunnel, 4.50° (IQR 6.34°); and POL tunnel, 4.69° (IQR 3.1°). No statistically significant differences were found in either the entry point or the angular deviation among the different bone tunnels. CONCLUSION: The use of 3D-printed PSI for lateral and medial anatomical knee reconstructions provides accurate and reproducible results and may be a promising tool for use in clinical practice.