Urban form, the physical characteristics of urban areas, can be used to predict environmental footprints, such as building energy consumption, carbon emissions, and water consumption. Building characteristics, such as height and footprint, are essential components of urban form models. Although progress has been made in predicting building characteristics to fill in observation gaps or derive 3-D architectural depictions, a gap in modeling efforts remains, particularly the relationship between building morphology and other landscape level patterns, such as land use and population.  These relationships have relevance for broader constraints on how cities evolve with landscapes in the future. A suite of models, the Building Morphology Distribution Land Model (BMDLM), was developed to determine how well building morphology can be predicted using land use (e.g. zoning) and population data at different resolutions. The model is showcased using Clark County, Nevada, and Los Angeles County, California as case studies. Random forest models were developed at two spatial granularities: 30m and 1-km resolution. Generally, 1-km models out-perform 30-m models, confirming higher resolution information is typically warranted. Models depicting building height and footprints as frequency distributions had the best model performance and lowest error, especially in Los Angeles County. Zoning and population were the most important predictors of building morphology of the 11 predictor variables tested. While modeling the distributions of building height and footprint with a machine-learning approach can yield reliable relationships between infrastructure and societal variables, predictive capabilities are overall poor (R² < 0.7).