Caching of popular content on wireless nodes is recently proposed as a means to reduce congestion in the backbone of cellular networks and to improve Quality of Service. From a network point of view, the goal is to offload as many users as possible from the backbone network to the wireless caches. However, if not done appropriately, this can lead to an unbalanced association of users to caches. As a result, some nodes can be overloaded while the resources of others are underused. Given a fixed content placement, this work proposes an efficient distributed algorithm to balance the association of user traffic among cellular cache memories. The algorithm finds the globally optimal solution and can be executed on base stations using a limited amount of information exchange between them. It is based on a novel algorithm we call Bucket-filling. The solution guarantees the utilization of each cache in a fair way rendering all available caches useful. The improvement compared to common user assignment policies is highlighted for single-as well as for multi-tier random networks.