An on-the-ground measuring device mounted on a tractor was used to measure crop cover percentage of wheat continuously in the field. An analogue video camera JVC TK-C1380 and an Imspector V9 spectrograph, together with a differential global positioning system (DGPS), were mounted on one end of the boom of the device. On 15 March and 27 April measurements were taken every 4 and 2 s for the camera and spectrograph respectively, while driving in the field. The field, consisting of 60 plots of 12 in by 16 in, had three different seeding densities and five different nitrogen application rates. On 8 May and at harvest on 10 August, crop samples were taken, measured and analysed in the laboratory to know crop characteristics. The crop variables were used as reference values for the goodness of the measured crop cover by the optical instruments. At the beginning of the growing season (15 March) seeding density influenced crop cover. Later (27 April), applied nitrogen affected the percentage crop cover. It was concluded that the spectrograph was more accurate in measuring crop cover percentage because the measurements were more related to seeding density in the beginning of the growing season (15 March) were more related to applied nitrogen midseason (27 April), were more related to crop measurements (biomass, nitrogen in the crop) midseason (8 May) and were more related to yield measurements at harvest on 10 August. (C) 2004 Silsoe Research Institute. All rights reserved Published by Elsevier Ltd.