Measured offsets of well-dated alluvial fan deposits near the Masonic Home in Union City constrain Holocene slip rate of the Hayward fault between 7 and 9 mm/yr. Our best minimum geologic slip rate over the past 8.27¿0.08 kyr (i.e., 8270 years) is 8.0¿0.7 mm/yr. A steep stream (its channel cut into bedrock) flows southwest out of the East Bay Hills, crosses the fault, and deposits its load on an alluvial fan. We cut two 5-m-deep, fault-parallel trenches 20--30 m southwest of the main fault through the crest of the fan. Walls of the trenches reveal a series of nested distributary channel fills. These channels had cut into old surfaces that are indicated by paleosols developed on flood silts. We distinguished many channel fills by their shape, clast size, flow direction, elevation, and relation to paleosols, enabling us to correlate them between both trenches. Two distinct episodes of fan deposition occurred during the Holocene. Reconstructing the apex positions of these fan units indicates that about 42¿6 m and 66¿6 m of fault slip has occurred since their inceptions at about 4.58¿0.05 ka, and 8.27¿0.05 ka, respectively. We lowered the age and age uncertainty of the younger unit from earlier reports based on new multiple radiocarbon dates. The 4.58 ka slip rate of 9.2¿1.3 mm/yr is not significantly different at 95% confidence from the 8.27 ka slip rate of 8.0¿0.7 mm/yr. Because current regional strain rates are fully consistent with Neogene plate tectonic rates (Lisowski et al., 1991) and the historic surface rate of creep in Union City is only 4.7¿0.1 mm/yr (Galehouse, 1994), the larger, ≥8 mm/yr, Holocene slip rate implies that strain is now accumulating on a locked zone at depth. The 8 mm/yr rate is probably minimal because earlier trenching evidence nearby implies that some unknown additional amount of fault deformation occurs outside of the narrow fault zone assumed in measuring slip. Lienkaemper et al. (1991) suggest that the fast creep rate of 9 mm/yr, measured near the southern end of the Hayward fault, may underestimate the deep slip rate, because 1868 surface slip occurred there in addition to the continuing fast creep. If the historic deep slip rate equals the long-term rate, then the 9 mm/yr creep rate reflects the minimum seismic loading rate of the Hayward fault better than the ≥8 mm/yr Holocene rates do. ¿ American Geophysical Union 1996