Objective: The aim of the present study was to investigate the association of one-year change in organizational justice (i.e., procedural justice and interactional justice) with job performance in Japanese employees. Methods: This study surveyed 425 men and 683 women from a manufacturing company in Japan. Self-administered questionnaires, including the Organizational Justice Questionnaire (OJQ), the World Health Organization Health and Work Performance Questionnaire (WHO-HPQ) and the scales on demographic characteristics, were administered at baseline (August 2009). At one-year follow-up (August 2010), the OJQ and WHO-HPQ were used again to assess organizational justice and job performance. The change in organizational justice was measured by dichotomizing each OJQ subscale score by median at baseline and follow-up, and the participants were classified into four groups (i.e., stable low, adverse change, favorable change and stable high). Analysis of covariance (ANCOVA) was employed. Results: After adjusting for demographic and occupational characteristics and job performance at baseline, the groups classified based on the change in procedural justice differed significantly in job performance at follow-up (ANCOVA: F [3, 1097]=4.35, p <0.01). Multiple comparisons revealed that the stable high procedural justice group had significantly higher job performance at follow-up compared with the stable low procedural justice group. The groups classified based on change in interactional justice did not differ significantly in job performance at follow-up ( p >0.05). Conclusions: The present findings suggest that keeping the level of procedural justice high predicts higher levels of job performance, whereas the psychosocial factor of interactional justice is not so important for predicting job performance.(J Occup Health 2015; 57: 388–393)