Hail is a major cause of crop loss and property damage in the United States. Hailstreaks are columns of hail that have swept the ground. Devegetation of the land surface by hailstreaks can have significant biogeophysical consequences. Changes in the surface energy balance and local wind fields can give rise to “inland sea-breeze” phenomenon that may trigger convection and severe weather. We investigated the relationship between hail occurrences and the appearance and persistence of hailstreaks in time series of satellite images. Due to abrupt changes in vegetation density, hailstreaks can be identified in Normalized Difference Vegetation Index (NDVI) imagery. To detect more clearly changes in vegetation, difference images were generated from a standard set of biweekly maximum AVHRR NDVI composites for the conterminous US produced by the USGS EROS Data Center. These image data have a nominal spatial resolution of 1 km. Overlaying the digitized point locations of the National Weather Service reports of hail onto the NDVI difference images, hailstreaks were identified as dark areas coincident with or proximate to hail reports. From 1990-1999, 118 events of significant hailstreaks were observed. The hailstreaks ranged in length from 9 to 367 km (median=61.3 km) and in area from 21 to 8443 sq km (median=407.5 sq km). Hailstreak persistence was a complex function of seasonal timing of the event, vegetation type and phenology, and event severity.