Background: The original 32-item Auditory Hallucination Assessment Scale (AHAS) was translated into Chinese by our research team which had been used to measure auditory hallucination for schizophrenia patients. However, it was suggested that the lengthy questionnaire should be modified to reduce participants' burden of answering questions and a simple form should fit better to psychiatric and research settings. Based upon the advice, we have tried critical ratio to eliminate some questions from the initial lengthy form of Auditory Hallucination Assessment Scale (AHAS-LCF). The purpose of the study was thus aimed to develop a simple Chinese form for Auditory Hallucination Assessment Scale (AHAS-SCF) and to further enable it a reliable and valid test for schizophrenia patients in Taiwan. Methods: The participants in research were 209 schizophrenia hospitalized patients at the average age of 41.89 from Northern Taiwan. We adopted a stepwise regression with forward procedure to select testing items. The dependent variables were the overall scores on the lengthy-form of AHAS (AHAS-LCF). The independent variables were the scores for each item in the AHAS. All of the testing items were included in regression models in which a forward stepwise procedure was used to identify the best predictors of the total scores. We repeated the procedures and analyzed their internal consistency until both of AHAS-LCF and AHAS-SCF displayed equivalent. The factor analysis then proceeded to identify the domains of AHAS. The internal consistency of AHAS-SCF was evaluated by using Cronbach's α., and the test-retest reliability was evaluated by intra-class correlations and Spearmen correlations. Results: The overall score of short form AHAS and its dimensions were correlated with the long form's (all r > 0.951, p < 0.001). The EFA turned out three domains with eigenvalues >1, explaining 37.40, 11.90 and 7.75% of the variance, respectively, accounting for a total of 57.11% of the variance. The three domains were given names respectively: 'characteristics' (6 items), 'beliefs' (3 items), and 'emotional-behaviors' (10 items). The total 19 items on the AHAS-SCF showed internal consistency (Cronbach's α) >.820; the test-retest reliability run by Intra-Class Correlation (ICC) was .98, and the Spearmen Correlation .96 for the total AHAS score. Conclusions: The research results proved that the AHAS-SCF can be properly used for assessment of schizophrenia patients in Taiwan.