While measurements of polar meteorology and ocean currents have received much attention in recent years, surface radiation data have lagged considerably. This is despite their importance to calculations of the surface heat budget, cloud radiative forcing, and albedo. Those radiation time series which do exist are preponderantly coastal or located on stable platforms, but data return is limited to seasonal information about local changes in short- and longwave radiation. The few datasets which have sampled regional and basin-scale radiation from floating ice stations and shipboard cruises report radiation values over a larger expanse but not necessarily over a more varied surface. For example, ice stations are necessarily located on multi-year floes which are less likely to melt through, while some ship operations are conducted entirely in open water or 10/10ths ice, depending on the scientific focus. Despite their inherent limitations, these combined data, both land and marine, are used to develop heat budget models for understanding air-sea-ice transfer and polar-optimized radiative parameterization schemes. In order to improve our understanding of radiative exchanges at the surface and to optimize the applicability of algorithms in models by specifying the uncertainties, measurements representative of the rapidly varying cryosphere is necessary. Within and along the boundaries of polynyas, such variations occur on scales easily sampled by ship and aircraft and appropriate to mesoscale modeling. In many cases, at least one of the polynya margins is bounded by land, where coastal-based data are relevant. Such a measurement scheme was employed during the North Water 1998 field season in the polynya occurring between Ellesmere Island, Canada and the Hayes Peninsula of Greenland. Simultaneous radiation data from well-calibrated sensors were recorded both at the Cape Herschel ice camp on the eastern polynya boundary and over the ice-ocean polynya surface from an ice-breaking research vessel during spring transition and summer melt. The time series from the two platforms were used to initialize a surface heat budget box model and test a number of clear-sky short- and longwave radiative parameterizations, which provided values for the calculation of cloud radiative forcing. Between the two sites substantial spatial variability existed, both in cloud cover and surface type, such that each radiation dataset was best represented by a different combination of short- and longwave schemes. As with all in situ polar data sets, these are likely to be dominated by regional effects. With larger, more extensive sets of measurements, it will be possible to determine more universal relationships between radiation and the changing polar environment. During the International Polar Year (2007-2008), the opportunity can be seized to expand upon the scant, current knowledge of surface radiation. Numerous aircraft and shipboard experiments, both planned and proposed, should incorporate radiative measurements. Particular preference for polynya measurements, where the data return is optimized by the variations in surface cover and proximity to instrumented coastal sites (some of which may require enhancements to their sensors), is stressed.