OBJECTIVES: The purpose of this study is to examine the measurement invariance of the WHOQOL-OLD Module across different demographic groups in Taiwan’s elderly people. By testing the differential item functioning (DIF), the presence of DIF indicates measurement variance. In other words, item score of measurement scale is dependent on some irrelevant characteristics. If this condition is observed the interpretation of items would be confounded.  METHODS: A total sample 524 participants involved in this study, but only 503 participants (96%) completed the whole questionnaire. The questionnaire included the WHOQOL-OLD and the demographic characteristics (age, gender, and education). Multiple-indicator multiple-cause latent variable (MIMIC) model was used to explore the relationship between the latent factors and demographic variables. Data was analyzed by using the Mplus7.0 and SAS9.4 software. RESULTS: Multiple-indicator multiple-cause latent variable model showed significant differences were observed on sensory abilities, death and dying for different age and education and on social participation only for different education. However, the significant differences can be eliminated when adjusting for DIF on item1 and item18, respectively, on sensory abilities and social participation. CONCLUSIONS: The result implies that age and education should be controlled when comparing sensory abilities, death and dying and social participation facets. Some items influenced by DIF might be another potential problem. Further studies on item effects are needed in reporting quality of life.