A laboratory experiment was performed to study a case in which a snow cover introduced on an established saline ice sheet resulted in physical processes that significantly affected the microwave brightness temperature over a period of a few weeks. Saline ice was grown to a thickness of 240 mm in an outdoor pool at ambient air temperatures. Precipitation was allowed by use of a movable roof. Brightness temperatures were measured at 10 and 85 GHz before and for several weeks after one snowfall. During the same period, the vertical temperature profile and crystallography of the snow column, as well as ice structure and salinity at the original ice surface, were monitored. The 10-GHz brightness temperature dropped by as much as 100 K from bare ice values during the first few days after the snow fell, because of a saline slush layer which formed at the bottom of the snow. The saline water in the slush layer apparently was forced up through the unbroken ice by the added snow load. The slush layer eventually froze into an added highly emissive frazil ice layer which raised the 10-GHz brightness temperature to above its bare ice values. The frazil ice layer was similar to superimposed frazil ice observed on freezing leads in high-latitude ice packs. The 85-GHz brightness temperature did not change from bare ice values soon after the snowfall but dropped by about 40 K over the following 20 days. We use a simple dielectric model to qualitatively test the dependence of 10-GHz brightness temperature on relevant physical conditions at the bottom of the snow. At 85 GHz the snow layer was optically thick, and the brightness temperature drop was probably the result of increased volume scatter from the growing snow grains.