STUDY DESIGN: Retrospective cohort study. OBJECTIVE: To investigate the diagnostic accuracy of lumbar movement restrictions and pain in patients with metastatic bone cancer. BACKGROUND: When evaluating patients with low back pain, physical therapists have used the presence of nonmechanical findings during a spinal movement screen as 1 factor in determining whether physician referral is necessary. There are no studies that have investigated the accuracy of this strategy in a situation of diagnostic uncertainty. METHODS: This study included 1,109 patients (655 women) with low back pain (mean age=54.8±16.3 years) seen at a spine surgery center who received a clinical movement screen and an imaging-supported diagnosis by an orthopedic surgeon. No report of pain during movement and no limitation of movement were considered the 2 targeted findings as these are associated with nonmechanical findings. RESULTS: Sixty-six patients were diagnosed with metastatic cancer, 61 with metastatic bone cancer and concomitant diagnoses. Pain-free lumbar movements in all directions for patients with metastatic bone cancer without concomitant diagnoses were associated with a posttest probability of 0.00 (+likelihood ratio=2.4; -likelihood ratio=0.0), which may be useful in ruling out spinal cancer. In situations where a concomitant diagnosis was present with cancer, the value of a movement screen was poor. CONCLUSION: Nonmechanical findings during a traditional movement screen are not specific to sinister conditions such as metastatic spinal cancer. Clinicians should expect concomitant conditions to exhibit painful or limited findings in patients with and without cancer.