Aim Happiness is an important indicator of mental and physical health. It has been emphasized as one kind of well‐being, and its definition varies from culture to culture. The main objective of the present study was to examine the psychometric integrity and dimensions of the Chinese Happiness Inventory (CHI) in relation to scores on Ryff's Psychological Well‐Being Scale among retired older people in Taiwan. Methods A cross‐sectional study was carried out at social service centers in Taipei, Taiwan. Retired adults gave informed consent from September to November 2010, and completed a package of structured questionnaires measuring happiness and psychological well‐being. Internal consistency, the factor structure of the CHI and criterion validity were assessed. Results Results from an exploratory factor analysis showed a three‐factor solution for the CHI. These factors were named Positive Affect, Life Satisfaction and Interpersonal Relationships. Internal consistency coefficients were 0.95 (Positive Affect), 0.91 (Life Satisfaction), 0.85 (Interpersonal Relationships) and 0.97 (total scale). The results of a canonical correlation analysis showed the presence of a strong relationship between CHI and Ryff's Psychological Well‐Being Scale (r = 0.69), and that two canonical variates could be derived from the relationship between them. Conclusions The results show that the CHI is a three‐dimensional scale with high reliability and validity. The construct of happiness emphasizes relationships in relation to others and environment rather than autonomy in this sample. Although the components of happiness might be similar for Positive Affect and Life Satisfaction, their weights for Interpersonal Relationships should be considered when measuring happiness in different cultures. Geriatr Gerontol Int 2016; 16: 865–872.