This study was designed to estimate the prevalence and diagnostic validity of Schneider’s first-rank symptoms of schizophrenia in a sample of Arab patients with a major mental illness. One hundred sixty-eight (n=168) patients were selected for the purpose of this study from a larger population of patients who were previously recruited in the clinical field trials of the ICD-10 and in studies related to the reliability of the Arabic-English version of the Schedules for Assessment in Neuropsychiatry (SCAN). Most of the patients were assessed by pairs of clinicians, and in over two thirds of the cases diagnostic structured interviews were applied. The OPCRIT system was applied to these patients to yield diagnostic categories along different operational diagnostic criteria of schizophrenia (DSM-111, DSM-111-R, ICD-10, RDC, Feighner’s, Schneider’s and French criteria). Frequencies of individual first rank symptoms were found to be low but discriminated significantly among psychotic patients with the exception of delusional perception. Their sensitivity and specificity ranged between 0,4 to 0,6 and 0,75 to 0,89 respectively, depending on the operational definition of schizophreniSensitivity of Schneider’s symptoms was highest among patients with RDC and ICD-10 schizophreniEstimates of the diagnostic efficiency of Schneider ‘s symptoms suggest that they add significantly to the accuracey in the diagnostic classification of schizophreniThe authors conclude that by deleting Schneider’s FRS from the diagnostic algorithm of schizophrenia, it is highly unkikely that improvement in the classification rate of the disorder will take place in the absence of pathognomonic symptoms that are yet to be ascertained.