In this paper, we present two parallelization schemes for solving a particular class of plate-bending problems using the finite strip method on a network of workstations equipped with the well-known software package PVM (parallel virtual machine). The first scheme is based on the master-slave model and the second is a multilevel master-slave approach. The finite strip method, first developed in the context of thin plate bending analysis, is a semi-analytical finite element process. It takes advantage of the orthogonality properties of harmonic functions in the stiffness matrix formulation to yield a block diagonal stiffness matrix and, thus, decomposes a single problem in two dimensions into several independent subproblems, each corresponding to a problem in one dimension. The method not only reduces the bandwidth of the system stiffness matrix, but offers obvious advantages for solving the problem on either shared-memory machines or distributed-memory multiprocessors. Our parallelization strategies are implemented on a network of SUN workstations interconnected by the Ethernet. Numerical experiments and the performance of the implementations are presented. Our results clearly indicate that the multilevel master-slave model is better than the master-slave scheme on the networked system using PVM.