The emission of solar flares at millimeter wavelengths is of great interest both in its own right and because it is generated by the energetic electrons which also emit gamma rays. Since high-resolution imaging at gamma-ray energies is not presently possible, millimeter observations can act as a substitute. Except for that class of flares known as gamma-ray flares the millimetric emission is optically thin. It can be used as a powerful diagnostic of the energy distribution of electrons in solar flares and its evolution, and of the magnetic field. We have carried out high-spatial-resolution millimeter observations of solar flares this year using the Berkeley-Illinois-Maryland Array (BIMA), and report on the preliminary results in this paper (Kundu et al 1990; White et al 1990). We also report some recent results obtained from multifrequency observations using the VLA (White et al 1990). M i l l i m e t e r B u r s t s Most millimeter-wave observations of solar flares have been hampered both by poor spatial resolution (because of the lack of synthesis interferometers) and by poor sensitivity (since the flux from the Sun's thermal emission is so high at millimeter wavelengths). The number of reported observations of bursts at millimeter wavelengths is relatively small, and there are none for which true imaging data have been reported. The BIMA array, presently consisting of three antennas but being expanded to six to produce an aperture synthesis imaging instrument with 15 baselines, has a spatial resolution of < 1" arc, and time resolution as good as 0.3 sec. In its present configuration there are too few baselines to image rapidly time-variable phenomena on the Sun, but the range of spacings available at BIMA (2" 60") is ideal for the study of flare sources; with data from three baselines we can determine the spatial scales involved and study the time evolution at high sensitivity and resolution. If good phase information is available, we can in principle obtain positions of burst sources.