Accurate estimates of wildland fire perimeters and areas are essential for planning wildfire response, monitoring prescribed fire, estimating pollution emissions, and for other natural resource applications. Remote sensing can provide a low-cost and relatively accurate means to monitor burned area on the landscape. The most common methods of remote sensing use the Normalized Burn Ratio (NBR), which is the ratio of reflectance bands sensitive to burned areas, or differenced Normalized Burn Ratio (dNBR), which is the difference between pre- and post-fire NBR. The NBR and dNBR methods each can have advantages in different situations. Reflectance values are categorized into levels of burn severity using field-measured values of the Composite Burn Index (CBI). However, these methods have not been calibrated for the dominant vegetation types of the southeastern United States. Our objective was to calibrate and test the accuracy of these methods for remotely measuring burn area. We established 731 CBI measurement plots in prescribed burned areas within the Apalachicola National Forest in Florida during the 2007 and 2008 dormant season (November-February), early growing season (March-June), and late growing season (July-October) in flatwood and upland (sandhill) pine (Pinus spp.) forests to determine NBR and dNBR breakpoints delineating burned versus unburned areas. We mapped the perimeters of selected burned areas on the ground using Global Positioning System (GPS). Corresponding burned areas were estimated using the NBR and dNBR methods with the newly determined breakpoints for comparison with surface-measured areas. The average percent bias in estimating burned area was -5% (± 15% SE) using NBR and -1% (± 7% SE) using dNBR and was not significant based on t-tests. However, the percent error of commission plus error of omission ranged from 4 to 92% (average 22%) using NBR and from 0 to 38% (average 14%) using dNBR. Percent error increased with time elapsed between the burn and the post-fire Landsat flyover, revealing time limit bounds for the accurate use of this method. Our findings suggest that NBR and dNBR imagery may provide an unbiased method for inexpensively monitoring burned area from fires > 10 ha in common southeastern U.S. habitats under the recommended set of conditions. © 2010, Tall Timbers Research, Inc.