Sieh and Jahns (1984) forecasted that the next moderate Parkfield earthquake might trigger a major earthquake along a fault segment greater than 30 km long southeast of Cholame. Their forecast assumed (1) the slip was 3-4 m in 1857 and characteristic of the segment; (2) a slip rate of 3.4 cm/yr; and (3) full strain release in earthquakes. This study represents an independent measurement of channel offsets, on 1:2400-scale low-sun aerial photographs and by field investigation, to estimate the amount of 1857 slip. Although rainfall is only moderate (30 cm/yr), few reliable offsets of less than 20 m persist here because cattle grazing and agricultural disking of soft sediments on the steep terrain greatly aggravate erosion. Reconstruction of offset geometry and size depends heavily on assumptions made about the post-1857 erosion. Most of the apparent 3- to 4-m offsets of Sieh and Jahns (1984) can also be measured as 2 to 3 m larger with equal or lower uncertainty. The four offsets judged as most reliable range between 5.4 and 6.7 m, and the 11 offsets of medium-high reliability average 5.8 ?? 0.3 m. Data are too sparse and ambiguous to resolve details of the 1857 slip for this segment but it is distinctly less than the 9 m of the Carrizo Plain and more than the 3-4 m previously estimated. Further trenching may refine some measurements, but probability calculations for a Cholame segment earthquake must allow for large observer-dependent uncertainty in the 1857 slip. Although the probability of an M ???7 Cholame event seems less than that suggested by a 3.5-m characteristic earthquake model, it remains among the highest in the state.