$$O^{18}/O^{16}$$ ratios of 550 specimens of snow, firn, and ice predominantly from West Antarctica have been determined. Identification of annual layers in Antarctic snow and firn by O-isotope variations is undoubtedly complicated by wind action. Nonetheless, $$O^{18}/O^{16}$$-ratio curves with seemingly normal seasonal characteristics obtained at the South Pole, Byrd, Little America V, and Wilkes satellite stations consistently suggest accumulation rates 20-100 per cent greater than determined by surface measurements and pit stratigraphy. Comparative annual accumulation values in cm. depth of water (values derived from $$O^{18}/O^{16}$$ determinations in parentheses) are as follows: South Pole, 7 (15); Byrd, 14-18 (36); Little America V, 15.5-24 (30); and Wilkes satellite, 13 (15). The $$O^{18}/O^{16}$$ data are consistent enough and the variations in values large enough to support the suggestion that mean annual rates of accumulation in West Antarctica may have been underestimated. New-fallen snow at Ellsworth station shows the usual direct variation in $$\delta^{3}$$ of the $$O^{18}/O^{16}$$ ratio with temperature over an impressively large range. Samples taken within individual snowfalls display different $$\delta$$/temperature relationships depending upon the nature and history of the air masses involved. Samples of wind-drifted snow from Little America V have a seasonal flavor, but their $$\delta$$ values are considerably lower than those of snow from a nearby pit. A speculative explanation of this anomaly is that some of the wind-driven snow comes from inland sources, possibly as much as 500 km. away. Since this low-$$\delta$$ material does not seem to accumulate permanently at Little America V, consideration must be given to the alternative explanation that wind action produces mechanical separation favoring accumulation of ice particles with higher $$\delta$$ values. The low $$\delta$$ material may end up in the ocean. Quantitative relationships between variations in $$\delta$$ and surface air temperature, upper-air temperature, and tropospheric temperature are not overly impressive. Lack of consistency in the $$\delta$$/surface-air temperature relationship can be attributed to the strong inversion frequently prevailing in Antarctica, but even the influence of upper-air and tropospheric temperatures appears to be variable. However, lower temperatures at any level are always associated with more negative $$\delta$$ values. This suggests that factors other than temperature are involved such as the previous history of the air mass, the length of overland travel and the height to which the air has had to ascend. For example, the more negative mean $$\delta$$ value at Ellsworth station compared to Little America V may be partly due to the fact that some of the air masses reaching Ellsworth station crossed the polar plateau. Estimates based on the $$\delta$$/temperature relationships suggest that material at a depth of about 290 meters at Byrd station accumulated under temperature conditions 2°-4° C. colder than now prevailing.