General objectives for measuring solar ultraviolet (UV) irradiance are: a) To establish a UV climatology by long‐term monitoring, e.g. within a network, b) To detect trends in global UV irradiance, c) To provide datasets for specific process studies and for the validation of radiative transfer models and/or satellite derived UV irradiance at the Earth’s surface, d) To understand geographic differences in global UV irradiance, e) To gain information about actual UV levels and their diurnal and seasonal variability, f) To provide data for public information and awareness (e.g. UV index). Solar ultraviolet radiation can be measured by different classes of instrument and it is crucial to match the instrument employed to the intended objective. It should also be ensured that sufficient facilities are available to support the objective in terms of personnel and QA/QC requirements. The three classes of instrument available for solar UV measurement are spectral, broadband and multifilter. Spectral instruments are the most costly, complex and demanding of those available, but provide the most versatile data. Broadband radiometers, tend to be cheaper and have fewer operational problems than spectroradiometers. However, their maintenance and QA/QC can introduce substantial additional cost. Multifilter radiometers combine some of the properties of both broadband and spectral instruments. Not all the above mentioned instruments are suited to all of the objectives. To achieve objectives a,d,e and f, any class of instrument can be used, but it is necessary that it accomplishes a minimum requirement in quality. The second listed objective, trend detection, is the most demanding goal of UV monitoring, and spectral instruments are most suited to this task. The instrument specifications and the QA/QC requirements necessary to enable detection of small trends are very stringent and must be maintained over a prolonged period to justify trend detection. Providing datasets for process studies or satellite validation (objective c) is again best served by spectral instruments. For the validation of radiative transfer models the accuracy of spectral measurements must be comparable to the accuracy needed for trend detection.