Benford's law applies to a wide variety of natural and man-made datasets and it has been successfully applied to test the integrity of data and to detect possible fraud and anomalous results in economy, politics, and management studies. In this paper, we investigate the applicability of Benford's law to lightning data. To do this, we use lightning data in Switzerland obtained using the European Cooperation for Lightning Detection (EUCLID) network. The considered dataset consists of the cloud-to-ground flashes in Switzerland for the period from 1999 till 2007. First the total number of negative and positive flashes per day is considered in the analysis. It is shown that the obtained distribution is in very good agreement with Benford's law. The same analysis is repeated considering in the dataset only the flashes containing return strokes with absolute peak currents lower than 5 kA, for which the detection efficiency of the lightning location network is expected to be lower. The resulting distribution shows less agreement with Benford's law. The obtained results suggest that Benford's law may find an interesting application in the evaluation of the detection efficiency of a given lightning location network.