Vectors use a data model in which data, usually categorical or discrete data, is stored and represented visually as coordinate points, or vertices. A single vertex is a point; multiple vertices strung together can form lines (referred to as arcs in ESRI products), polygons, or even three-dimensional objects in virtual space; these vector-based objects can be rasterized with relative ease.

GIS software takes advantage of the characteristics of vector data to generate maps in which the features on the map are defined by vertices. Each vertex is georeferenced, often taking advantage of Global Positioning System (GPS) satellites; each feature on the map is then described by attributes, which provide the user with information that can be used to locate features and perform geospatial analysis.

Vector datasets are flexible, and their file size is usually small: they grow in size only as more data is added to the dataset. The primary disadvantage of vector datasets is their inability to represent phenomena without distinct boundaries.

Compare: rasters.