Mantle convection patterns of the past are not well known, yet an understanding of changing mantle convection characteristics is fundamental to understanding the evolution of plate tectonics. There are very few ways to examine mantle characteristics of the past. Changes in spreading rate and volcanic activity with time have been used to draw comclusions about historic changes in mantle activity. Mantle temperature has been found to be related to crustal thickness. With this relationship, crustal thicknesses may now yield new conclusions about historic changes in mantle characteristics. We have inferred changes in mantle convection patterns throughout the last 180 m.y. by examining variations in assumed crustal thickness within the Pacific basin. Crustal thicknesses were calculated from residual depth anomalies by assuming that residual depth anomalies are the result of isostatic compensation of variations in crustal thickness. Crustal thickness is determined at the time of crustal formation and is dependent upon the temperature of the mantle source material. Intraplate hot spot volcanism effects on crustal thickness were not ignored. Examination of variations in crustal thickness of crust of different ages can reveal information about changing temperatures of the mantle at the ridge through time. We have learned that mantle temperatures at the ridge during the mid-Cretaceous were more variable than those temperatures at the ridge after the mid-Cretaceous. Furthermore, we have inferred from the data that mantle temperatures at hot spots were higher during the mid-Cretaceous than those at hot spots existing after the mid-Cretaceous. We suggest that mantle convection at the ridge was more rapid during the mid-Cretaceous causing a higher variability of temperatures at the ridge. We also note that this period of increased mantle convection is concurrent with the increased mantle temperatures at hot spots within the Pacific basin. ¿ American Geophysical Union 1996