In this paper we have proposed Text Summarization using genetic algorithm. According to the required summary size by the user, we have evaluated all possible subsets of that size using genetic algorithm. We have made an attempt to create a summary that is less redundant and covers all the topics in the document. The summary generated was evaluated by comparing it with the user summary and the corresponding rouge score was calculated. In the end, the influence of varying summary size is analyzed and is compared with the human summary.