When evaluating the performance of a photovoltaic (PV) system, it is extremely important to correctly measure the plant operating conditions: incident irradiation and cell temperature. At large-size PV plants, the possible dispersion of the plant operating conditions may affect the representativeness of the values measured at one single point. The available literature contains many observations on irradiance dispersion (typically associated to high temporal resolution experiments) and its effects on the PV power output (unexpected power transients, power fluctuations, etc.). However no studies have been made on the long-term energy-related effects of geographic dispersion of solar irradiation, which could affect, for example, to the uncertainty in determining energy performance indexes like PR. This paper analyses the geographical dispersion in the PV operating conditions observed at low temporal resolutions (day, month and year) at two PV plants located, respectively, in the south of Portugal and the north of Spain. It shows that daily irradiation deviations are significantly higher than is commonly supposed. Furthermore, once the measurement points are a certain distance apart (a few hundred metres), the deviations in irradiation appear to be independent of distance. This could help to determine how many irradiance sensors to install in order to reduce uncertainty. Daily mean temperature differences between different points at a large-scale PV plant range from 1 to 7 K and are not related to the distance between measurement points.