Alfalfa accounts for the largest cropping area in both the High Desert and Intermountain regions in California, and the use of site-specific management (SSM) can potentially improve farmers’ fertilization practices and crop nutritional status. These areas have limited to no studies regarding nutrient SSM, and variable rate (VR) fertilizer application has not been commonly used by farmers in either area. Considerable range of soil nutrient levels have been indentified in some alfalfa fields in California, however, this variability has not been considered for nutrient management. The objectives of this project were to compare three soil sampling densities (A = 1 sample/1.2 ha, B = 1 sample/2.4 ha and C = 1 sample/4.8 ha) in order to establish a pattern for future soil sampling for nutrient variability assessment in alfalfa fields in the High Desert and Intermountain regions, and to compare fertilizer usage and cost differences between uniform rate (UR) and VR application methods. Two hundred and four samples were collected in five alfalfa fields located in Lancaster and near Yreka, CA, based on a sampling grid density of 1 sample/1.2 ha. Most of the soil phosphorus (P) and potassium (K) variability and fertilizer savings due to VR occurred in the fields located in the Intermountain region. Overall, maps created based on sampling densities A and B were very similar. Total fertilizer savings due to VR application in all 5 fields (247.3 ha) combined was $3,823. Although VR application resulted in the use of 12% more K fertilizer in a particular field of 36ha located in the Antelope Valley region, application rates on that field ranged from 56 to 280 kg ha^-1. This fact emphasizes that the more intensive soil sampling of the VR method allowed the identification of portions of fields where soil K would be overestimated with the UR method.