Background: The purpose of this study was to investigate outcomes of chronic patients unresponsive to previous spinal manipulative therapy subsequently treated with manipulation under anesthesia (MUA). Methods: A prospective outcome cohort study was performed on 30 patients who had not improved with previous treatment and who underwent a single MUA by a doctor of chiropractic. The numeric rating scale for pain (NRS) and Bournemouth Questionnaire (BQ) were collected at 2 weeks and 1 day before MUA. At 2 and 4 weeks after MUA, the Patient's Global Impression of Change, NRS, and BQ were collected. The intraclass correlation coefficient evaluated stability before treatment. Percentage of patients "improved" was calculated at 2 and 4 weeks. Wilcoxon test compared pretreatment NRS and BQ scores with posttreatment scores. Mann-Whitney U test compared individual questions on the BQ between improved and not improved patients. Logistic regression compared BQ questions to "improvement." Results: Good stability of NRS and BQ scores before MUA (intraclass correlation coefficient=0.46-0.95) was found. At 2 weeks, 52% of the patients reported improvement with 45.5% improved at 4 weeks. Significant reductions in NRS scores at 4 weeks (P=.01) and BQ scores at 2 (P=.008) and 4 weeks (P=.001) were reported. Anxiety/stress levels were significantly different at 2 and 4 weeks between improved and not improved patients (P=.007). None of the BQ questions were predictive of improvement. Conclusion: Approximately half of patients previously unresponsive to conservative treatment reported clinically relevant improvement at 2 and 4 weeks post-MUA.