With regard to rapid development of Internet technology and the increasing volume of data and information, the need for systems that can guide users toward their desired items and services may be felt more than ever. Recommender systems, as one of these systems, are one of information filtering systems predicting the items that may be more interesting for user within a large set of items on the basis of user’s interests. Collaborative filtering, as one of the most successful techniques in recommender systems, offers some suggestions to users on the basis of similarities in behavioral and functional patterns of users showing similar preferences and behavioral patterns with current user. Since collaborative filtering recommendations are based on similarity of users or items, all data should be compared with each other in order to calculate this similarity. Due to large amount of data in dataset, too much time is required for this calculation, and in these systems, scalability problem is observed. Therefore, in order to calculate the similarities between data easier and quicker and also to improve the scalability of dataset, it is better to group data, and each data should be compared with data in the same group. Clustering technique, as a model based method, is a promising way to improve the scalability of collaborative filtering by reducing the quest for the neighborhoods between clusters instead of using whole dataset. It recommends better and accurate recommendations to users. In this paper, by reviewing some recent approaches in which clustering has been used and applied to improve scalability , the effects of various kinds of clustering algorithms (partitional clustering such as hard and fuzzy, evolutionary based clustering such as genetic, memetic , ant colony and also hybrid methods)on increasing the quality and accuracy of recommendations have been examined.