Radon risk maps are usually based either on indoor radon data, or on measurements of soil gas radon and soil permeability. If these data are not available or not sufficient, it was suggested that other data could be used as an approximate substitute (a proxy) to the missing information, like the concentration of 238U or 226Ra in soils or the terrestrial gamma dose rate (TGDR). We examine here the correlation between airborne measurements of soil U and indoor radon, and between airborne U and TGDR, and their link with affected/unaffected areas. No clear correlation is found between airborne U and affected areas, as strongly affected areas are not characterised by a higher U level. Only the moderately affected area of Condroz can be connected to a higher U level, related to a few U anomalies. TGDR shows a rather good correlation with airborne U, but its relation with radon risk is less clear. Soil uranium and TGDR may help to screen out areas with very low U and very low TGDR, which have a low indoor radon risk, but they cannot be considered as good proxies for predicting radon affected areas in Belgium.