An important part of the Energy Balance Experiment (EBEX-2000) was the measurement of the net radiation and its components. Since the terrain, an irrigated cotton field, could not be considered homogeneous, radiation measurements were made at nine sites using a variety of radiation instruments, including pyranometers, pyrgeometers and net radiometers. At several of these sites multiple instruments were employed, which enabled us to compare instruments and assess accuracies. At all sites the outgoing longwave and shortwave radiation and the net radiation were measured, while the incoming radiation was supposed to be uniformly distributed over the field and was therefore measured at three sites only. Net radiation was calculated for all sites from the sum of its four components, and compared with the direct measurement of net radiometers. The main conclusions were: (a) the outgoing shortwave radiation showed differences of up to 30 W m −2 over the field; the differences were not clearly related to the irrigation events; (b) the outgoing longwave radiation showed differences of up to 50 W m −2 ; the differences increased during the periods of irrigation; (c) the net radiation showed differences of several tens of W m −2 across the field, rising to 50 W m −2 or more during the periods of irrigation; (d) the net radiation is preferably to be inferred from its four components, rather than measured directly, and (e) attention should be paid to the characteristics of pyranometers that measure the outgoing radiation, and thus are mounted upside down, while they are commonly calibrated in the upward position. The error in the net radiation at EBEX-2000 is estimated at max (25 W m −2 , 5%) per site during the day and 10 W m −2 at night.