In this paper, a new and powerful algorithm called Flower Pollination Algorithm (FPA) is proposed for optimal allocations and sizing of capacitors in various distribution systems. First the most candidate buses for installing capacitors are suggested using Power Loss Index (PLI). Then the proposed FPA is employed to deduce the size of capacitors and their locations from the elected buses. The objective function is designed to reduce the total cost and consequently to increase the net saving per year. The proposed algorithm is tested on 15, 69 and 118-bus radial distribution systems. The obtained results via the proposed algorithm are compared with other algorithms like Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Plant Growth Simulation Algorithm (PGSA), Direct Search Algorithm (DSA), Teaching Learning-Based Optimization (TLBO), Cuckoo Search Algorithm (CSA), Artificial Bee Colony (ABC) and Harmony Search Algorithm (HSA) to highlight the benefits of the proposed algorithm. Moreover, the results are introduced to verify the effectiveness of the suggested algorithm to minimize the losses and total cost and to enhance the voltage profile and net saving for various distribution systems.