A new survey of radio sources at 151 MHz, which has not been described previously, is in progress at Cambridge. There are several of us working on it including Warner, Kenderdine, Waggett, Masson and Mayer. The results of the first observations are at present in a preliminary state but we hope that in time they will form the first part of the 6C survey. The purpose of the survey is not to reach the faintest sources detected so far in aperture synthesis observations but to study moderately faint sources at a low observing frequency and to cover a large part of the northern sky rapidly. The deepest survey made so far at a low frequency is that of Ryle and Neville (1962) at 178 MHz over a region of 50 square degrees near the north celestial pole. The faintest sources detected had flux densities of 0.25 Jy, corresponding to a source density of 104 sr−1. It is already 15 years since that survey, which was the first trial of aperture synthesis using the earth's rotation, and much more is now technically possible. One of the most interesting features of a low frequency survey is its ability to detect preferentially sources with steep radio spectra and to be sensitive to sources of very low surface brightness. We know that in many cases these two properties go together and are associated with old radio sources, or at least with those parts of sources which are old. Many of the weak radio galaxies in nearby clusters are obvious examples of this type of source while the final, and so far unidentified, stages of the development of the most powerful double sources may be exciting candidates for discovery.