Sprite halos were recently identified as an impulsive but spatially diffuse phenomenon that sometimes occurs just prior to, but distinct from, sprites. The lack of discernible spatial structure and the temporal development sequence in halos differs markedly from the highly structured bodies and tendrils and the complex development sequences of sprites. However, both phenomena are thought to result from an electric field due to charge moment changes usually associated with large positive cloud-to-ground (CG) lightning but also following negative CG flashes. Three-dimensional triangulations of sprites and sprite halos were made between stations in South Dakota and Wyoming in August 1999 during the NASA Sprites99 balloon campaign. Halos were found to have a Gaussian 1/e diameter of ~66 km and 1/e thickness of ~4 km. Comparison with the location of the underlying lightning strokes, as recorded by the National Lightning Detection Network (NLDN), confirms that the horizontal position of sprites may be laterally offset by as much as 50 km from the underlying parent lightning discharge, as has been previously reported. The point of maximum apparent brightness for sprite halos occurs at an altitude of ~78 km, similar to that of sprites. However, unlike sprites, this point tends to be centered directly above the underlying parent lightning discharge, 4.6¿2.7 km mean distance from the center of the halo to the NLDN location. This difference in spatial location relative to the underlying lightning suggests that the electrical breakdown associated with discrete sprites may require a random ionizing event such as a micrometeor. In contrast, sprite halos do not appear to require such a random component. ¿ 2001 American Geophysical Union