Abstract: We studied the usefulness of remote sensing to predict aphid density in wheat (Triticum aestivum L.) fields in Oklahoma. Remote sensing and aphid data were collected in three fields on three consecutive days starting on December 17, and ending December 19, 2003. A Cropscan multispectral ground radiometer and an Ocean Optics hyperspectral walking spectrometer were used to gather reflectance data over a total of 24, 0.25 m² wheat plots per field. Greenbug, Schizaphis graminum (Rondani,) and birdcherry oat aphid, Rhopalosiphum padi (L.), were counted in each 0.25 m² plot. Twenty five spectral vegetation indices were calculated from the reflectance data and four regression techniques: linear, non-linear, robust MM, and robust least trimmed squares (LTS) were performed to investigate the best relationships among the aphid densities and spectral vegetation indices. Variability in spectral vegetation indices explained by total aphid densities (greenbug + bird-cherry oat aphid) were 75%, 58%, and 75% for the data collected in fields one, two, and three, respectively. The proportion of variation in vegetation indices attributable to greenbug densities were 61%, 69%, 73% for the data gathered in fields one, two, and three, respectively. Birdcherry oat aphid densities explained 46%, 45%, and 75% of the variation in spectral vegetation indices for fields one, two, and three, respectively. These results were obtained from the second degree non-linear regression models suggesting that remotely sensed data can be used to estimate aphid densities in wheat fields with acceptable degree of accuracy.