Technical Abstract: Synthetic weather generators are important for continuous-simulation of agricultural watersheds for risk analyses of downstream water quality. Many watersheds are sparsely or totally ungauged and daily weather must either be transposed or augmented. Since water quality models must recognize runoff, erosion, sediment transport, and pollutant loads, augmented precipitation statistics must reflect the precipitation patterns as well as the average annual precipitation of the indigenous parent’s historical climate. This paper addresses how well synthetic weather generators reproduce their parent’s historical weather’s precipitation amounts and patterns which are to be used for subsequent simulation of watershed runoff, erosion, streamflow and pollutant loads. Synthetically generated climate can create time series of daily weather that are long enough to conduct risk analyses; i.e., frequency and probability analyses of the resulting streamflow and pollutant loads. Historical records frequently are serially incomplete and generally too short, or simply may not be available, for such purposes.