This study aims to enrich the comprehension of the effect of living density on anxiety and stress among adults in a global city. A random sample of 1,978 Hong Kong adults was interviewed in a cross‐sectional population study in 2014–2015. Descriptive statistics and logistic regressions were used to investigate the association between housing variables and mental health indicators, namely, anxiety and stress. Logistic regression analysis results have shown that by controlling the confounding effect of demographic variables, income poverty, housing ownership and housing cost, the living density was significantly associated with anxiety and stress of residents. Compared with those living in high density of <7 m2, living in medium density of ≥7 and <13 m2 was significantly associated with lower risk of anxiety (adjusted OR 0.52, 95% CI 0.30–0.88), and the risk was less for those living in low density of ≥13 m2 (adjusted OR 0.41, 95% CI 0.23–0.72). Meanwhile, living in low density of ≥13 m2 was significantly associated with a lower risk of stress (adjusted OR 0.44, 95% CI 0.20–0.97). These results highlighted the significant impact of living density on personal anxiety and stress. Moreover, female, younger adults or those living in income poverty were also at risk of anxiety and stress. In conclusion, our findings implied that improving housing policies, such as building public housing with adequate living areas and market regulation of living density of private housing, would help enhance the mental well‐being of residents.