A cloud-sensing Doppler radar is used with a vertically pointing antenna to measure the vertical air motion in clouds during the Atlantic Stratocumulus Transition Experiment. The droplet fall velocity contamination was made negligible by using only measurements during the time the reflectivity was below − 17 dBZ. During one day of measurements, the daytime character of the vertical velocity variance is different than that of the nighttime case. In the upper part of the cloud, the variance had a distinct maximum for both day and night; however, the nighttime maximum was about twice as large as the daytime case. Lower down in the cloud, there was a second maximum, with the daytime variance larger than the nighttime case. The skewness of the vertical velocity was negative near cloud top in both the day and night cases, changing to positive skewness in the lower part of the cloud. This behavior near cloud top indicates that the upper part of the cloud is behaving like an upside-down convective boundary layer, with the downdrafts smaller in area and more intense than the updrafts. In the lower part of the cloud, the behavior of the motion is more like a conventional convective boundary layer, with the updrafts smaller and more intense than the downdrafts. The upside-down convective forcing in the upper part of the cloud is due to radiative cooling, with the daytime forcing less because of shortwave warming.