Objectives:The purpose of this study was to evaluate the characteristics and the effects of neurofeedback from child and adolescent patients with neurosis. Methods:Thirty-two patients who received neurofeecback from hospital in Korea from July, 2005 to July, 2010 participated in this study. The demographic data, characteristics of psychiatric disorders, and states of neurofeecback were analyzed. CGI(Clinical Global Impression) score and The Hill-Castro Checklist was evaluated before and after neurofeedback training. Results:Depression was the most common psychiatric disorder(16 patients, 50%), followed by anxiety disorder(10 patients, 31.3%). The remainder of the study group included patient suffering from adjustment disorder, somatoform disorder. Thirty patients(93.8%) has been taking medicine, and fifteen patients(46.9%) of those patients has been taking medicine for less than a year. The average frequency of neurofeecback was 33.84, and neurofeedback protocol was individualized according to patient’s symptom. In this study, neurosis patients preferred α/θ(53.1%) training. Significant change of CGI after training was noticed using covariance with frequency(<.001), and self rating scale also showed significant changes in inattention, hyperactivity, impulsivity, immaturity, depression, self esteem and hostility(<.001). Conclusion:This is the retrospective study in clinical setting, so there are several limitations. But this study demonstrates the significant positive effects of neurofeecback in objective and subjective rating scales for child and adolescent patients with psychiatric disorders. Prospective controlled studies are needed in the future.