In order to test the validity of the Magsat anomaly data, the recently published U.S. Geological Survey/Society of Exploration Geophysicists Composite Magnetic Anomaly Map (CMAM) of the conterminous United States was used as a standard, and a comparison was made with these satellite data. The CMAM data set, digitized by the Phoenix Corporation, was resampled and continued upward to 320 km altitude. This upward continued data set was compared with an independently derived Magsat field at the same altitude. Our initial results were unsatisfactory; the CMAM data dominated by a long-wavelength surface increasing in amplitude by some 400 nT in a northwesterly direction across the United States. This broad trend surface was suppressed by removing a general two-dimensional polynominal function having nine orders in longitude and five in latitude obtained through least squares fitting. The resulting residual upward continued field was compared with the Magsat data by means of correlation techniques. Correspondence was found between thse data sets for many large-amplitude anomalies, for example, off the East Coast, in Kentucky, Oklahoma, Mississippi embayment, Georgia, Arizona, and Nebraska. Dissimilarities occur over the Rio Grande rift, Minnesota, and Wyoming. The negative anomaly over the Rio Grande rift and the positive anomaly over the midcontinent geophysical anomaly are more pronounced in the CMAM data than in the Magsat results. These latter discrepancies are attributed to the method used to remove external field variations from the satellite data which suppresses north-south trending structures, i.e., those lying in the orbit direction. CMAM base level problems localized over Minnesota and Wyoming are believed to cause observed discordance between satellite and aeromagnetic data in these regions. This study confirms, with ''ground truth'', the validity of these satellite measurements and demonstrates that satellite-derived anomaly maps can be used to identify base level problems when compiling local or regional aeromagnetic surveys.