This paper describes the parallelization of a mesoscale-cloud-scale numerical weather prediction model and experiments conducted to assess its performance. The model used is the Advanced Regional Prediction System (ARPS), a limited-area nonhydrostatic model suitable for cloud-scale and mesoscale studies. Because models such as ARPS are usually memory and CPU bound, the motivation here is to decrease the computer time required for running the model and/or increase the size of the problem that can be run. A domain decomposition strategy using a network of workstations produced a significant decrease in elapsed time and increase in problem size relative to a single-workstation run. The performance of the resulting program is described by deprived formulas (collectively known as a performance model), which predict the execution time and speedup for different numbers of processors and problem sizes. The interprocessor communication speeds are shown to be the major obstacle to achieving full processor use. The effect of faster communication networks on parallel performance is predicted based on this performance model. Parallelization experiments using the ARPS code were run on a cluster of IBM RS6000 workstations connected via Ethernet. The message-passing paradigm implemented here made use of the library of routines from the Parallel Virtual Machine software package.