Because 3He is produced by spallation of C, O, and Ca, calcite may be a suitable target for cosmogenic 3He dating. Samples of crystalline calcite veins in limestone cobbles were collected from three inset fill terraces in eastern Tajikistan (~3800 m), which are offset by the Karasu fault. The upper three terrace levels have 56 ± 1.2, 40 ± 1.9, and 18 ± 0.6 Mat/g of 3He (n=3-5 samples for each level). We believe this 3He is cosmogenic in origin because: 1) rocks from a given terrace level yield very similar concentrations, 2) these concentrations are consistent with relative terrace ages, 3) 3He/4He ratios of 20-350 Ra far exceed ratios expected for trapped or radiogenic components, and 4) Li contents in the calcite are extremely low (< 50 ppb) so nucleogenic 3He is negligible. To evaluate the potential for diffusive loss of 3He from these calcites, two samples were proton irradiated to make large quantities of 3He and subjected to step-heating diffusion experiments. Resultant Arrhenius plots show a lightning bolt pattern in which successive prograde or retrograde heating steps are linearly related in (1/T) vs. ln(D/a2) space, but ln(D/a2) decreases for successive isothermal steps. Based on a mean annual temperature of ~3° C, and modeling following Wolf et al. (1998), we conclude that for 30,000 yr time scales, these calcites should be >99 % retentive to cosmogenic 3He. The ages of the terraces are presently unknown; assuming a reasonable cosmogenic 3He production rate of 130 at/g/yr gives ages of 24, 18, and 7.5 Ky, corresponding to nearly identical strike-slip rates of ~2.75 mm/yr for both of the upper two terraces. These preliminary results suggest that cosmogenic 3He in calcite may be a valuable dating tool in young samples found in cool climates but possibly not in hot climates.