A good understanding of soil spatial variability within and between fields can help improve productivity in terms of overall crop yield, quality and evenness; and the use of nutrients and appropriate cultivations. This chapter sets out the main options for determining the variability of soil properties and how this can be developed into effective soil and land management strategies. There are a range of field survey approaches, technologies and statistical methods that can be used in isolation or in combination to map variation in soil properties; from soil samples analysed in the laboratory, soil survey (using a spade, auger and an experienced soil surveyor) to proximal sensing (e.g. electromagnetic induction) and geostatistical survey. The most appropriate and cost-effective method will depend on the amount of prior information on soil variability, the local cost of available skills and technologies (e.g. the cost of soil analysis), the scale of operation, and the specific objectives of the survey (i.e. how does the grower intend to use the spatial information to vary their land management?). Where strong spatial variability in soil physical properties is expected (e.g. variability in sand, silt, clay or organic matter content) but there is no prior spatial information on that variability, the use of proximal sensing may be useful as a means of guiding soil sampling schemes and establishing soil management zones. Geostatistical surveys are especially useful where weak spatial variability is expected in soil physical properties (e.g. soils formed from widespread uniform sandy or silty deposits) but strong spatial variability is expected in soil chemical properties, such as soil nutrient reserves, which may have arisen due to contrasting management between or within fields. In this context, a spatial covering design with additional close pair points can yield useful information but the overall costs of sampling and analysis should be borne in mind.