External fixation pins tend to loosen with time, especially from cancellous bone. A coating that releases a bisphosphonate has been shown to improve the fixation of dental implants in humans. We now tested a bisphosphonate coating on steel pins for external fixation. The primary hypothesis was that coated pins would be better fixed in the diaphysis. 20 patients with medial knee osteoarthritis underwent proximal tibial correction osteotomy with hemicallotasis. They received a pair of pins (Orthofix) in the tibial shaft, one bisphosphonate-coated and one uncoated. Another pair of pins was inserted in the metaphysis, near the joint. This pair was a bisphosphonate-coated pin and an HA-coated pin. All pins were inserted according to a random list. The pins were removed after the osteotomy had healed (8-15 weeks), and extraction torque served as the predetermined main outcome variable. No pins showed clinical signs of loosening. Removal torque for the shaft pins was 6.6 Nm (SD 2.2) for the bisphosphonate and 6.0 Nm (SD 2.5) for the uncoated (difference = 0.5, 95% CI: -0.03 to 1.3). Removal torque for the metaphyseal pins was 4.4 Nm (SD 1.3) for the bisphosphonate-coated and 4.2 Nm (SD 1.6) for the HA-coated (difference = 0.2, 95% CI: -0.5 to 1.0). We could not show any improved cortical fixation, but the metaphyseal findings are striking. In a previous study on 19 patients with a similar layout, HA-coated and uncoated pins were compared. In the metaphysis, all 19 uncoated pins loosened before removal. It was concluded that uncoated pins could not be used in the metaphyseal region. The present results suggest that a bisphosphonate coating enables metaphyseal fixation similar to that of hydroxyapatite coatings, with no difference from uncoated pins in cortical bone.