Supplying a proportion of the N requirement of a wheat crop via the foliage would potentially reduce immobilization of fertilizer N in the soil organic matter and N losses by leaching or denitrification. A field experiment was carried out at Harper Adams in Shropshire to investigate the effect on crop yield of supplying the spring N application to winter wheat as different proportions of urea as a solution rather than as conventional soil-applied urea, and to determine the physiological basis of any yield differences. A solid ammonium nitrate treatment was included to represent alternative commercial practice to solid urea. Treatments were repeated on the same plots over the 3 years 1992, 1993 and 1994. Solid fertilizer was applied as a single dressing, whereas urea sprays were split over a number of days to reduce scorch. Nitrogen as urea sprays produced similar grain yields to N applied conventionally to the soil as solid ammonium nitrate or urea, but effects on above-ground dry matter production and harvest index depended on the time of application. Application of a large proportion of N as urea sprays, such that some of the N as urea solution was applied later in relation to crop development, produced less above-ground dry matter, but compensated by increasing harvest index. It is concluded that application of N as urea sprays could be successfully used to substitute for soil-applied N fertilizer at stem extension in winter wheat without loss of yield. Extra application costs, however, are likely to outweigh any efficiency or environmental benefits, except where applications of solid N are made to dry soils.