The multi-angle imaging spectroradiometer (MISR) will be reporting band-weighted spectral radiances for its level 1B1 radiometric product. Although the out-of-band is a small percentage of the in-band response, the entire range from 365 to 1100 nm can potentially contribute to the reported radiance. For certain geophysical parameter retrieval algorithms, therefore, it is desirable to remove the out-of- band contribution. This will be done for select level 2 products. In order to provide such a correction, some estimation must be made of the out-of-band scene spectral properties, so that this signal might be removed. For MISR this can be done by making use of information from all four bands (nominally 443, 555, 670, and 865 nm). This paper evaluates the effectiveness of a four point, piecewise- linear, approximation to the surface reflectance. This representation is derived from the four MISR bands. To evaluate this approach, data from the airborne visible/infrared imaging spectrometer (AVIRIS) is utilized. This sensor measures the total upwelling radiance from 400 to 2450 nm in the electromagnetic spectrum through 224 channels at 10-nm intervals. This study shows that, for MISR, there can be as large as a 5% difference in the total band-weighted spectral radiance, as compared to the desired in-band weighted spectral radiance. The MISR retrieved four point surface profile is sufficient to provide out-of-band correction to within 1% accuracy.