Soil freezing has been reported both as beneficial and detrimental to soil structure, but the subsequent tha« ing process has not been adequately investigated as a factor in determining the net effect of freezing and thawing] To assess the role of liquid water during thawing, hand-sieved naturally occurring 1-to 2-mm aggregates from three soils (0.11,0.33, and 0.44 kg kg"' day contents) at three initial water contents (0.10,0.20, or 0.30 kg kg"') were subjected to three treatments, including not frozen (control), freeze-thaw (freezing at -15°C for 12 h and thawing at 15°C for 12 h) and freeze-drying (overnight freezing at -15°C followed by further freezing at -éO°C and s ow thawing to room temperature under vacuum to promote sublimation). Aggregate stability was measured via wet agrégate stability and dispersible clay following treatment. The freeze-thaw treatment was typically destructive in nature while the freeze-drying treatment resulted in greater aggregate stability. The extreme desiccation of the aggregates through ice crystal growth within interaggregate pore spaces and subsequent sublimation are thought to be responsible. Conditions favoring sublimation and commonly experienced during the winter include bare soil surfaces and synoptic meteorological conditions of dear skies, low humidity, and moderate winds. Aggregate stability measured in the spring may therefore reflect the effects of drying of the soil aggregates via the freezing pro ;ess and the resulting pore water content distribution following thawing.Abbreviations: DC, dispersible clay; WAS, wet aggregate stability.