Two field experiments were conducted, one each with the energy grasses switchgrass, Panicum virgatum L. and reed canary grass, Phalaris arundinacea L., to determine the role of nutritional elements in optimizing crop yield and associated effects on crop quality for thermal energy conversion technologies. The two experiments were adjacent and on low productivity sandy soils in the South East of England. This paper reports the crop yield effects during the years 2003-2005. Nitrogen was added in 50 kg ha (-1) increments from 0 to 250 kg ha (-1) Nitrogen. The Potassium and Sulphur treatments were applied in combination with 50 and 150 kg ha (1) Nitrogen. No yield increase was detected, suggesting that the soil Potassium and Sulphur supply was already sufficient. In reed canary grass a yield response was recorded when 50 kg ha(-1) Nitrogen was applied. However, the greater statistical power achieved by pooling the Potassium and Sulphur treatments at the 2 rates of Nitrogen showed that application of 150 kg ha (1) Nitrogen compared to 50 kg ha (1) Nitrogen, to reed canary grass, increased crop yield in each of the three years of the study, and by a maximum of 1.35 t ha(-1) Dry Matter in 2004. The same comparison in switchgrass showed a lesser response (0.7 t ha(-1) Dry Matter) to 150 compared to 50 kg ha(-1) Nitrogen and in one year only (2003). Therefore, the data generally show reed canary grass to be the more responsive crop to nitrogen, but all data should be viewed in respect of the limitations to yield imposed by site conditions. For both crops, yield peaked in either late summer or early autumn and then declined overwinter, most markedly in 2004-05. Overall, mean crop yield at final harvest for both crops, in late winter of each year, was low (<4 t ha(-1) Dry Matter). This was considered to be due to water stress. The sandy soils had a low water holding capacity and the weather during 2003-05 was generally warmer and dryer than the long term average for the site. Block 3 of the reed canary grass experiment rooted into an area of finer textured soil (at ca. 1.5 m depth) and produced greater yields (5.45 cf. 3.25 and 3.01 t ha (-1) Dry Matter) from Blocks 1 and 2 respectively in 2004-05. The results clearly indicate that these genotypes of switchgrass and reed canary grass may fail to give financially viable yields on inherently low productivity sites, and that varieties specially selected for such conditions will need to be bred for the industry if such areas are to be used. (C) 2012 Elsevier Ltd. All rights reserved.