The identification of noun compound as multi-word lexical units is very important task in natural language processing applications that require some degree of semantic interpretation such as,machine translation,information retrieval and text summarization. In this paper,we used the hybird method for extracting the noun compound from arabic corpus that is based on linguistic knowledge and statistical measures. For the candidate indentification,we have used some linguistic analysis tools such as lemmatization and POS in order to filter the candidates and determine the variations. The association measures have been computed for each candidate to rank the candidates. After that,we have evaluated the association measures by using the n-best evaluation method. We reported the precision values for each association measure in each n-best list. The experimental results showed that the log-likelihood ratio is the best assocation measure that achieved highest precision. Keywords: Hybrid method, arabic noun compunt, association measures, morphological variations, lemmatization, n-best evaluation method.