OBJECTIVES: To describe a program to translate evidence into practice for the use of manipulation with a sub-group of patients with low back pain and report the program's outcomes following implementation. We compared outcomes based on appropriate inclusion in the program and compliance with the evidence being translated. METHODS: The evidence translation program was based on evidence that patients meeting two criteria (duration of symptoms <16 days, no symptoms distal to knee) were likely to respond to a physical therapy that included manipulation in the first two visits. Implementation addressed potential barriers with referring physicians, physical therapists, and scheduling staff to this evidence. Outcomes for patients in the program were tracked following implementation. Process outcomes were appropriateness of inclusion (met both criteria), compliance with evidence for providing thrust manipulation in the first two visits, and number of physical therapy visits. Clinical outcomes were based on Oswestry scores from the first, interim (after two to three visits), and final visit. RESULTS: A total of 577 patients entered the evidence translation program (mean age = 43.0, 56.8% female); 79.5% were appropriate inclusions and 83.0% received manipulation. The use of manipulation was associated with fewer visits (mean difference = 0.54 visits, 95% CI: 0.037, 1.04, P = 0.035), and appropriate inclusion was associated with greater Oswestry change (mean difference at the final visit = 6.6 points, 95% CI: 1.6, 11.6; P = 0.010). DISCUSSION: Implementing evidence into practice is difficult; however, barriers can be anticipated and overcome. Tracking the outcomes of an implementation program is critical to evaluating its benefit to patients. Additional research using experimental designs are necessary to evaluate the effectiveness of various treatments implemented in physical therapy practice.