The Ottawa Knee Rule (OKR) is a clinical decision rule for the ordering of knee radiographs by physicians in patients with blunt knee injuries. However, in many Emergency Departments, radiographs are also ordered by nurses during triage. This study was designed to compare application of the OKR by triage nurses and physicians. A consecutive sample of all patients presenting to the Emergency Department (ED) of a tertiary care military teaching hospital with blunt knee injuries that met OKR inclusion criteria were enrolled. Radiographs were ordered on all patients. Nurses and physicians were blinded to each other's examinations. Nurse-physician inter-reviewer reliability was compared using the kappa statistic. Of 172 enrolled subjects, 38 were excluded for incomplete data or multiple entries, leaving 134 study patients. Four fractures (3%) were detected. Fair inter-observer reliability was found on four of the five OKR components: patellar tenderness (PT, kappa = 0.31), proximal fibula tenderness (PFT, kappa = 0.31), inability to flex > 90 degrees (FLEX, kappa = 0.24) and inability to walk four steps (WALK, kappa = 0.44). Nurses were more likely to overestimate the presence of individual components (PT 3.5x, PFT 6.6x, FLEX 1.9x and WALK 4.8x) and were 3.6 times more likely than Emergency Physicians (EP), overall, to order radiographs, representing a 21% vs. 37% reduced radiography rate, respectively. No fractures were missed by physicians or nurses. Triage nurses and EPs in this study had fair agreement in their application of the OKR. Triage nurses greatly overestimated knee injuries, while maintaining sensitivity, at the expense of specificity and cost savings.