The importance of Antarctic coastal polynyas for heat exchange between ocean and atmosphere, for high ice production, and thus, with the resultant brine rejection, for a large amount of the Antarctic bottom water is widely recognized. To obtain full understanding of their influence, continuous measurement of even small polynyas is necessary. Only passive microwave sensors, with their global coverage and their ability to penetrate cloud cover, can provide this information. However, because of their coarse resolution, a special method to estimate the area of subpixel-scale coastal polynyas has been developed. It uses 85- and 37-GHz data successively in order to take full advantage of the higher resolution at 85 GHz while compensating for its sensitivity to atmospheric effects with the 37-GHz data. This method is based on simulating microwave images of polynya events by convolving an assumed brightness temperature distribution with the satellie antenna pattern. These images are compared with measured microwave data and the polynya area iteratively modified until best agreement is found. Application of the method to synthetic images produces a maximum error of 200 km2 but a mean error of 80 km2. The method shows distinct improvement over the more traditional, i.e., ice concentration, methods. Analysis of coincident infrared data indicate that ice with a thickness up to 0.06 m is included in the estimated open water area. Area time series derived with the method for a coastal polynya near Halley Bay show day-to-day changes from 50 km2 to 450 km2 during austral spring. Interpretation of these results in terms of wind forcing and ice growth in polynyas is aided by comparison with a one-dimensional model of polynya development. ¿ American Geophysical Union 1995