OBJECTIVES: To examine the factor invariance of the WHOQOL-OLD across gender, age, and resident area for the old people in Taiwan. METHODS: Data were collected from 512 Taiwan elderly people (M = 76.2, SD = 7.5, ages ranged from 60 to 99) including 177 males and 335 females. 232 aged from 60 to 75 and 279 aged from 76 to 99. 334 live in the southern rural area and 176 live in the northern metropolitan area. To examine the factorial invariance of the WHOQOL-OLD, the sample was divided into two groups on different gender, age, and resident area. First, a baseline six-factor model was tested for different gender, age, and resident area respectively. Second, multi-sample analysis was conducted across gender, resident area, and age. Specifically, equal constrains on factor loadings, error variances, and factor variances were imposed. Model comparisons by using Chi-square difference tests were conducted to examine the factor invariance across gender, age, and resident area. RESULTS: Multi-sample analysis revealed that the factor loadings were invariant across different gender, age, and resident area groups respectively. Besides, when imposing equal constrains on the factor loadings, item variances and factor variances, the model fit indices revealed that the only complete factor invariance model was across gender but not age and resident area groups. CONCLUSIONS: This study suggests the underlying factor construct of the WHOQOL-OLD are similar to different degree across different gender, age, and resident area groups.  We conclude that the WHOQOL-OLD is a practical measurement tool for different gender, age, and location groups for the old people in Taiwan.