Objectives Hikikomori is a social condition characterized by social withdrawal. It is essential to develop an objective indicator to screen for this condition and provide early intervention. Therefore, this study verified the validity and reliability of a Korean version of the 25-item Hikikomori Questionnaire (HQ-25). Methods Semi-structured interviews were conducted with participants at high risk of feeling loneliness and who were recruited through online and offline methods. Data based on clinical scales were collected, and after four weeks, a self-report questionnaire was administered to a few participants to confirm the degree of agreement. Results One hundred forty-three people were enrolled, out of which 45 were identified as hikikomori. There was a significant difference in scores between the hikikomori group and the unaffected group (64.44 and 39.30, respectively; p<0.01). Convergent diversity and test-retest reliability were satisfactory. The area under the receiver operating characteristic curve was 0.817. Based on confirmatory factor analysis, the existing model was found suitable. In this study, sensitivity and specificity were 88.9% and 52.0% at 40.5 points, and 84.4% and 49.0% at 41.5 points, respectively. Conclusion The Korean version of the HQ-25 has secure validity and reliability. This scale can be beneficial in the screening and early intervention of hikikomori.