A dense network of rainfall stations over a certain area gives results in neighbouring stations that are highly correlated. To avoid the establishment of a too dense network, evaluation of the properties of a network in a given area has to be carried out. For this reason the interstation correlation-distance relationship was analysed for a network of rainfall stations in the eastern part of the Netherlands. Two models are used to fit the plot for each month. Both models are of the negative exponential type, the first one with distances taken linear, the second one with distances taken quadratic. Extrapolation of each model to zero distance does not result in a correlation coefficient equal to 1. Discrepancies from this value are considered to represent all exposure variations of the instrument and of the environment as well. Better results than those obtained by extrapolating the model to zero distance cannot be expected with the same devices in the same area. The efficiency of the network must be given in a relative sense, eliminating the disturbing influences of all exposure variations.