Macropores created by biological or physical processes can profoundly influence water movement through the soil. In cold regions, natural processes such as wetting/drying and freezing/thawing can influence macropore stability. Little has been documented, however, concerning the over winter stability of macropores. This study measured changes in the physical dimensions of artificial macropores, and the accompanying soil water content, from autumn to spring in the northern US Corn Belt. Macropores 30 cm deep having diameters of 5, 10, and 15 mm were created in a Barnes loam in October 1999 and October 2000. Soil water content was measured by neutron attenuation and Time Domain Reflectometry (TDR). After snowmelt the following spring, all of the artificial pores were at least partially filled with soil. In the spring all of the 5- and 10-mm pores and more than 95% of the 15-mm pores were undetectable at the soil surface. The extent of the smaller pores maintained over winter was generally greater than that of the larger pores. Over both winters the 5-mm pores maintained approximately 65 to 70% of their created length, and the 10-mm pores maintained approximately 50% of their created length. The length maintained by the 15-mm pores varied greatly over the two winters, 46% over the 1999–2000 winter and 4% over the 2000–2001 winter. The difference in the stability of the 15-mm pores over the two winters was attributed to the much wetter conditions of the 2000–2001 winter. No significant difference in the soil water content before or after freeze-up was measured as the result of the presence or size of the macropores. The results of this study show that macropores are not stable over winter, but the collapse of the macropores had no clear effect on soil water dynamics the following spring.