New, high-precision U-series ages of solitary corals (Balanophyllia elegans) coupled with molluscan faunal data from marine terraces on the Pacific Coast of North America yield information about the timing and warmth of the last interglacial sea-level highstand. Balanophyllia elegans takes up U in isotopic equilibrium with seawater during growth and shortly after death. Corals from the second terrace on San Clemente Island (offshore southern California), the third terrace on Punta Banda (on the Pacific Coast of northern Baja California), and the Discovery Point Formation on Isla de Guadalupe (in the Pacific Ocean offshore Baja California) date to the peak of the last interglacial period and have U-series ages ranging from ca. 123 to 114 ka. The first terrace on Punta Banda has corals with ages ranging from ca. 83 to 80 ka, which corresponds to a sea-level highstand formed in the late last interglacial period. U-series analyses of corals from the Cayucos terrace (central California) and the Nestor terrace at Point Loma (southern California) show that these fossils have evidence of open-system history, similar to what has been reported by other workers for the same localities. Nevertheless, a model of continuous, secondary U and Th uptake shows that two ages of corals are likely present at these localities, representing the ca. 105 and ca. 120 ka sea-level highstands reported elsewhere. U-series ages of last interglacial corals from the Pacific Coast overlap with, but are on average younger than the ages of corals from Barbados, the Bahamas, and Hawaii. This age difference is explained by the nature of the geomorphic response to sea-level change: fringing or barrier reefs on low-latitude coastlines have an accretionary growth style that keeps pace with rising sea level, whether on a tectonically rising or stable coastline. In contrast, midlatitude, high-energy coastlines are sites of platform cutting during the early part of a sea-level high stand and terrace scouring and concomitant sediment and fossil deposition as sea level starts to recede. The youngest ages of corals from the Pacific Coast suggest that sea level was still relatively high at ca. 116 ka, which is not in agreement with other estimates of relatively large global ice volume at that time. Reliably dated, ca. 120 ka marine-terrace deposits on the Pacific Coast have fossil mollusks that indicate water temperatures as warm or warmer than at present. In contrast, ca. 80 ka marine deposits reported here and elsewhere have fossil mollusks indicating cooler-than-modern water temperatures. The presence of both ca. 105 ka and ca. 120 ka corals on the Nestor and Cayucos terraces explains a previously enigmatic mixture of warm-water and cool-water mollusks. At ca. 105 ka, a relatively high sea level with cool waters may have "captured" the terrace formed during the 120 ka sea-level highstand, in areas of low uplift rate. The inference of cooler-than-modern waters off the Pacific Coast of North America at ca. 80 ka and ca. 105 ka, based on marine-terrace faunas, does not agree with estimates of sea-surface temperatures derived from alkenone studies in the Santa Barbara Basin. However, cooler water temperatures at these times are in agreement with paleo-temperature estimates from planktonic foraminiferal data for the Santa Barbara Basin. All records, from central California to Baja California, whether from marine terraces or offshore cores, indicate at least seasonably warmer-than-modern waters during the peak of the last interglacial period at ca. 120 ka.