Question Answer System (QAS) has many methods in determining candidate answer and must have the right answer for each question. Previous QAS using Fuzzy logic focused on candidate and ranking of answer. However, the QAS needs improvement in determining the relevant answer from the question and generating the correct answer in the question answer process. In this paper, we propose a new method to combine fuzzy logic and retrieval passages to obtain a collection of relevant answers in order to obtain high answer accuracy. We take relevant answers from the collection of the answer document and choose the exact answer using fuzzy logic. Methods for the QAS are preprocessing, question analyzer, passage retrieval, passage scoring, scoring for similar text, measuring keyword and candidate answers, fuzzy logic controller, rules, and extraction answer. This study produced significantly relevant answers compared to the TF-IDF method. The performance of the system has improved the accuracy of QAS by 80% which is better than the previous study.