ABSTRACT The literature lacks a rigorous psychometric evaluation of the Arabic version of Hospital Anxiety and Depression Scale (HADS) in hemodialysis (HD) patients. This study aims to evaluate reliability, determine the underlying factor structure of the Arabic version of HADS and assess its suitability as screening tool for depression and anxiety among Arabic HD patients. A sample of 370 HD patients were recruited from all health districts in Kuwait. Reliability for HADS (all items) and its subscales HADS-A (anxiety) and HADS-D (depression) were estimated using Cronbach’s alpha and item analysis was conducted. Exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were conducted to extract and test the factor structure for the Arabic version of HADS. Eight models were tested using CFA to determine goodness-of-fit. The Cronbach α for the Arabic HADS (all items), HADS-A and HADS-D were 0.884, 0.852 and 0.764, respectively. Pearson correlation coefficient between HADS-A and HADS-D subscales indicated significant correlation (r = +0.69, PV < 0.001). EFA indicated two factors with eigenvalues >1, which accounted for 48.5% of the total variance. CFA revealed the one-factor model had the poorest fit, the two-factor models with acceptable fit, and three-factor models showed good fit. The Arabic HADS has good reliability and internal consistency, which warrants its use in screening for anxiety and depression among Arabic HD patients. Furthermore, the three-factor structure has shown a better fit which warrants further exploration in HD patients with the advancement in theory of psychological models for anxiety and depression.