The purpose of this study was to determine the variability of femoral tunnel positions applying two different lateral extra-articular tenodesis (LET) techniques, guiding on the tactile identification (1) of the lateral epicondyle (Lemaire procedure) and (2) of the Kaplan fibre attachments on the distal femur (MacIntosh procedure) and to analyse whether one of these procedures is more suitable for reliable femoral tunnel positioning in LET procedures. Two experienced knee surgeons determined femoral tunnel positions in ten fresh-frozen cadaveric knee specimens according to the modified Lemaire and MacIntosh techniques. Tunnel positions were measured on true lateral radiographs as absolute distances from the posterior cortex line (anterior-posterior direction) and from a perpendicular line intersecting the contact of the posterior femoral condyle (proximal-distal direction), respectively. Further, tunnel positions were measured relatively to the femur height. An independent F test was used to compare variances between Lemaire and MacIntosh tunnel positions and between anterior-posterior and proximal-distal directions. The mean Lemaire and MacIntosh positions were found 2.7 ± 5.5mm and 3.6 ± 3.4mm anterior to the posterior cortex line, and 7.5 ± 5.0mm and 17.3 ± 6.9mm proximal to the perpendicular line intersecting the contact of the posterior femoral condyle, respectively. Mean Lemaire and MacIntosh positions, relative to the femur height, were found at 8.8% and 10.9% (anterior-posterior) and 22.2% and 50.8% (proximal-distal), respectively. Variability in tunnel positioning was observed for both techniques, showing no significant differences in the variance (p > 0.05) and partially overlapping tunnel positions of both techniques. The overall variance of tunnel positions, however, was significantly greater in proximal-distal than in anterior-posterior direction (F = 2.673, p < 0.038). Femoral tunnel positions in LET procedures have a high degree of variability and inaccuracy. Palpating anatomic landmarks appears to be insufficient to generate reproducible tunnel positions. Radiographic landmarks may enable more reproducible identification of isometric femoral tunnel positions and reduce malpositioning.