The high radio positional accuracy now being achieved with long baseline interferometers, aperture-synthesis instruments and large filled-aperture telescopes makes the rapid and reliable processing of optical identifications essential. For some time now the customary approach for large surveys has been to carry out a search for possible radio-optical associations using the prints of the Palomar Sky Survey. Rough positions for optical counterparts have generally been estimated by interpolation within a small grid centred on the radio position. However, the final confirmation or rejection of the identifications has awaited spectroscopic or accurate positional data from one of the large optical telescopes, usually the 48-inch Schmidt or 200-inch reflector at Mt Palomar. It is becoming increasingly evident that not only is this process rather slow but it also means that preliminary interpretive work dealing with the identification content of a radio survey will be necessarily limited in scope.