The relief of Cretaceous guyots indicates that an enormous region of the Pacific plate was 3300-4800 m deep about 100 Ma ago. The plate, then 13-64 Ma old, had depth anomalies of -170 to -1600 m. It was drifting over a group of overlapping midplate swells that comprised a bumpy Darwin Rise of smaller extent than originally believed. In 100 Ma much of the region has subsided only 1000-1200 m instead of the 2000-3000 m predicted by thermal models. A post-Cretaceous sea level lowering of 200-500 m accounts for some of the difference. The remainder appears to require repeated thermal rejuvenation such as occurred in Tertiary time in the central South Pacific. The Jurassic and Early Cretaceous crust is not anomalously shallow just because it is old. It was shallow when it was young.