This paper presents some ideas based upon studies of historical data, intuition, and extensive synoptic experince-ideas which may be relevant to estimates of the climatological impact of a projected CO2 increase over future decades. Reliable 1000- to 700-mbar thickness data are used to demonstrate possible regional effects associated with global warming. Global thickness values are shown to be correlated with hemispheric mean temperatures given by several investigators, an it is believed that use of these thicknesses is more appropriate to large-scale circulation studies than is surface temperature. It is shown that the mean temperature of the entire contiguous United States is not highly correlated with the northern hemisphere mean. Precipitation over the United States is not well correlated with the northern hemisphere mean. Precipitation over the United States is not well correlated with global mean temperature either, although there is a suggestion that in summer, heavier precipitation occurs when hemispheric mean temperatures are low. Two scenarios for a moderate CO2-induced increase in temperature, a general hemispheric warming an a warming which is most intense in the Arctic, are examined by analysis of the warmest recent winters over the hemisphere and over the Arctic, respectively. Both cases are characterized by high-latitude blocking, relative cooling at lower latitudes, and strong regionalization of thickness anomaly patterns. A potential moderating influence by the oceans through increased CO2 absorption is discussed. Comparisons are made between latitudinal profiles of recent warm years an the results of the Manabe- Wetherald model studies. Overall, it appears that the climatic effects of CO2 variations may be much more complex than is often surmised, so it is not wise to draw inferences of regional anomalies from hemispheric means or latitudinal profiles.