We studied the spatial distribution, phenology of male display and daily activity patterns of the Eurasian Woodcock Scolopax rusticola in the central region of the Cantabrian Mountains (Asturias, Northern Spain), the southern limit of the breeding area of this species in mainland Europe. Displaying male counts were conducted during the breeding season in 2004 and 2009 and revealed roding activity from mid April to early July, with the highest display frequency from early May to mid June. We found clear altitudinal variation in the distribution of breeding Woodcock at the regional scale: roding males were not detected in our samples below 800 m above sea level (asl) (although we have evidence of Woodcock occasionally breeding at low altitudes) whereas the percentage of positive listening stations over 800 m asl exceeded 35%. Besides altitude, density of earthworms was the main predictor of Woodcock distribution at the landscape (mesohabitat) scale: earthworms were more abundant in woodlands where roding males were detected than in those where birds were not detected, and were also more abundant in woodlands than in nearby pastures, in contrast to the winter pattern reported in previous studies. Woodcock exhibited a slightly higher diurnal activity in summer, while in winter there was a much sharper contrast within the daily activity cycle and greater activity occurred at night. Microhabitat selection during summer was assessed by comparing vegetation structure at sites in which Woodcock were detected, either from radio-locations or by flushing birds out with pointing dogs, and at randomly selected points. Sites selected by Woodcock had a higher density of stems of less than 5 cm diameter, a lower density of stems of more than 40 cm diameter, and higher canopy cover at 0.5–2.0 m above ground than randomly selected sites. The vegetation structure of the habitat selected by Woodcocks would offer protection against predators and allow Woodcocks to move and search for food. The fact that structural and biotical determinants of Woodcock distribution at the two spatial scales considered were different may reflect a trade-off between competing demands manifested as scale-dependent association with habitat attributes. Based on the results of this study, we propose that the food (earthworm) abundance could be the main criterion for the Woodcock's habitat selection at the mesohabitat scale (landscape level), which would explain both the seasonal change in local distribution and in activity patterns, whereas the need for cover and protection from predators would be more important at the microhabitat scale (occupation of particular sites).