In 1967-96 Long-eared Owls in SW-Drenthc (20,000 ha) were censused twice a year during twilight and at night, i.e. in March-April (territorial activities) and May-July (mainly begging fledglings). At the same time, twelve study plots (totalling 3147 ha) within the confines of SW-Drenthe have been investigated more intensively by means of the combined mapping method. The population fluctuated between 18 and 128 pairs (Figure I depicts the distribution in the peak year of 1990), showing pronounced peaks and troughs. The long-term trend appeared to be either slightly increasing (overall pattern for SW-Drenthe), or stable/slightly decreasing (well-surveyed plots). The population is probably largely stable, because the mapping method in the intensively surveyed plots within SW-Drenthe is thought to have produced the more reliable data (Figure 2). The number of territories without begging fledglings showed smaller annual fluctuations during the census period (from II to 48 pairs, i.e. almost factor 5) than the number of successful pairs (from 5 to 95 pairs, i.e. factor 19). Pairs with fledglings showed a characteristic saw-tooth pattern in conjunction with outbreaks of common voles Microtus arvalis (Figure 3). Severe winter weather, as expressed bij Unsen’s frost index, was not correlated with numbers of Longeared Owls in the following breeding season (total number of territories for SW-Drenthe: Spearman, rs=0.176, P>0.05), Flowever, prolonged snowfall may have some impact, as witnessed after the snow-rich winter of 1978/79. Apparently, the overall pattern of peaks and troughs is largely the outcome of fluctuations in the number of successful pairs. It remains to be investigated whether this signifies more pairs attempting to breed in vole-rich years, an influx of pairs following vole outbreaks, or both. Using standardised census techniques, monitoring of Long-eared Owls can reliably reveal ups and downs in large census areas. The census period should at least encompass two complete vole cycles, in order to compensate for short-term trends, i.e. a period of at least ten years in the case of SW-Drenthe (notice prolonged dips in numbers, as in 1982-88 and 1991-95; Appendix 1).