The aim of the study was to investigate test-retest reliability and construct validity of the World Mental Health Japan (WMHJ) version of World Health Organization Health and Performance Questionnaire (WHO-HPQ) short version according the COSMIN standard. We conducted two consecutive surveys of 102 full-time employees recruited through an Internet survey company in Japan, with a two-week interval in 2018. We calculated Pearson's correlation (r) of measures of the WHO-HPQ with other presenteeism scales (Stanford Presenteeism Scale, Work Functioning Impairment Scale, and perceived relative presenteeism), health and psychosocial job conditions. We tested the test-retest reliability (intraclass correlation, ICC) among those who reported no change of job performance during the follow-up. Among 92 (90%) respondents, the absolute presenteeism significantly correlated with WFun and perceived relative presenteeism (r=-0.341 and -0.343, respectively, p=0.001) and psychological distress (r=-0.247, p=0.018). The absolute/relative absenteeism did not significantly correlate with the other covariates. The test-retest reliability over a two-week period was high for the WHO-HPQ absolute presenteeism (ICC, 0.73), while those for absolute/relative absenteeism measures were moderate. The study found an adequate level of test-retest reliability, but limited support for the construct validity of the absolute presenteeism measure of the WMHJ version of the WHO-HPQ. Further research is needed to investigate the construct validity of the WHO-HPQ measures in a larger sample.