Buffer allocation is a challenging design problem in serial production lines that is often faced in the industry. Effective use of buffers (i.e. how much buffer storage to allow and where to place it) in production lines is important since buffers can have a great impact on the efficiency of the production line. Buffers reduce the blocking of the upstream station and the starvation of the downstream station. However, buffer storage is expensive both due to its direct cost and the increase of the work-in-process inventories it causes. Thus, there is a trade-off between performance and cost. This means that the optimal buffer capacity and the allocation of this capacity have to be determined by analysis. In this thesis, we focus on the optimal buffer allocation problem. We try to maximize the throughput of the serial production line by allocating the total fixed number of buffer slots among the buffer locations and in order to achieve this aim we introduced a new heuristic algorithm called “Line Balancing Algorithm (LIBA)”applicable to all types of production lines meaning that there is no restriction for the distributions of processing, failure and repair times of any machine, the disciplines such as blocking, failure etc. and the assumptions during the application of LIBA in the line.