BACKGROUND: The purpose of this study was to investigate the psychometric properties of the Arabic version of the CES-D. METHODS: Three samples of young Arab females were chosen from different young female populations (n =450). The translated version of the CES-D, sociodemographic and risk factors questionnaires were applied to the above three samples. A subsample was selected (n = 30) to re-rate the scale items and the same sample was interviewed by the SCID. The internal consistency and test-re-test reliability were estimated as well as the factor structure of the Arabic CES-D. The discriminative and criterion validity of the scale was also investigated using the ROC analysis. RESULTS: Seventy five percent (n = 350) of the subjects completed the scale. Alpha coefficient was found to be 0.88 and the split-half reliability was 0.83. The average re-test reliability (ICC = 0.59). Factor analysis yielded three factors: interpersonal problems, mixed affective/somatic factor and the positive affect factor. There was significant differences in mean total score between depressed and non-depressed subjects as classified by SCID as well as by self-perception of depression and the need for treatment. ROC analysis revealed that a cut-off point of 21 discriminated best between depressed and non-depressed subjects (sensitivity= 82%, specificity = 83% with AUC of 84%). CONCLUSION: The Arabic CES-D was found to possess reasonable reliability and reasonable discriminative and criterion validity.