The habitat, food and population dynamics of the field vole (Microtus agrestis) were studied mainly in southernmost Sweden (Scania) with a few comparisons from a northern locality. The habitats of Microtus agrestis consist mainly of successional stages, arising from abandoned cultivated fields, drained mires and felled mature forests. They were characterized botanically and classified according to moisture and time since abandonment. Capture distributions were significantly clumped, being most contagious at the lowest densities. Microdistribution was affected more by shelter than by food conditions. The weight of the stomach contents showed various skewed distributions, giving indications of the feeding behaviour of various population categories. The staple food in southernmost Sweden was various grasses, while herbs, gnss seeds, vegetative storage organs and moss were primary food items during summer, summer-autumn, winter and early summer respectively. In a northern locality herbs played a more important role in summer. Among plant species e.g. Agrostis spp., Festuca rubra, Achillea millefolium and Ranunculus spp. were often eaten, while the common Deschampsia caesjJiiosa and Charnaenerion angustifoliurn were disliked. There were clear differences between habitats but plant species belonging to the same group replaced each other in the diet. There were few differences between populatipn categories. Experiments disclosed the same preferences as found at stomach analysis. Digestibility was higher during spring than winter, that of the preferred species being highest. The digestibility was most affected by the proportions of crude protein but the changes in weight by the amounts consumed. The animals showed a decided interest in certain carbohydrates and fatty substances with differences between seasons and years. The interest in sugars was not correlated with the similar interest in bark. The probability of capture was especially low in juveniles and during winter. Trap lines and grids showed the same trappability, but, the intervals between checkings influenced the removal rate. The density changes on two distant, large, dry fields in southernmost Sweden were correlated, while a nearby isolated population in a spruce plantation fluctuated out of phase. The increases during the reproductive season were significantly lower than those expected at exponential increase. The density in various habitats did not depend on the moisture conditions. Differences in population structure and body weight in various areas were partly related to the phase of population development. However, there were also differences in the same area between habitats, which might differ in plant nutrient supply. The animals had a low energy reserve in their depot fat, which would only sustain the animals for 5-14 hours in an emergency. It was highest during midwinter but with significant depressions in April and December. Drier habitats could sustain populations of little more than 200 animals/ha over :winter and these densities were not realized in southernmost Sweden. The spring-summer density variations in a northern locality were correlated with the spruce seed supply during the following autumn-winter. A relation was assumed with the quantity or quality of previous storage tissues. In southernmost Sweden, with irregular and low-amplitude fluctuations, there was no clear relation to primary production.