Cross-validation is a popular computational method for model assessment and selection. With spatial data, however, many of the independence assumptions behind cross-validation break down. This talk will motivate and introduce some spatial cross-validation methods proposed in the literature to address these issues. We will then explore the results of a simple simulation study comparing the performances of nonspatial and spatial cross-validation methods on simulated spatial data. Though there is more work to be done, initial simulation results suggest that spatial methods indeed outperform nonspatial ones when applied to model assessment and selection in spatial contexts.