A population study of Terrapene carolina (Linnaeus) was made at the Patuxent Research Refuge, Maryland, from 1944 to 1947. A thirty acre area in bottomland forest was selected for intensive study. Turtles were marked by filing notches in marginal scutes according to a code. Turtles make extensive use of brushy shelter during the day as well.as at night. Gully banks and woods openings are used for sunning. Nights are usually spent in a 'form,' constructed by the turtle in leaves, debris, or earth. A form may be used once or it may be used repeatedly by the same or different turtles. Weather conditions most favorable to turtle activity are high humidity, warm sunny days, and frequent rains. Periods of activity are alternated with periods of quiet, even in favorable weather. There is no evidence for territorialism. Ranges of turtles of all ages and both sexes overlap grossly. Turtles are frequently found near each other but no antagonistic behavior has been observed. Adult turtles occupy specific home ranges which they maintain from year to year. Turtles retained their ranges even though a flood that completely covered the study area. Maximum home range diameters were determined by measurements of the mapped ranges of individual turtles. There was no significant difference between sizes of male and female ranges: males 33O+ 26 feet, females 37O+29 feet. A trail-laying device was used in following travel routes for 456 turtle days. Normal movements within the home range are characterized by (1) turns, doublings, detours, and criss-crossing paths, (2) interspersion of fairly direct traverses of the home range, (3) frequently repeated travels over certain routes. Maximum limits of the home range are ordinarily reached within a few days or weeks, although some turtles cover only one portion of the range at a time. Some turtles have two home ranges. One of these turtles was followed with a trailer for 161 days in 1946 and 1947. Trips outside the home range are made by females for egg laying. Trips of unexplained nature are made by both males and females. Turtles from other areas are occasional transients through the study plot. Standardized census trips provided data for estimating the size of the population. Census data were taken after females had returned from egg laying. The samples were spaced at intervals of a week or more to allow free movement of turtles over their ranges and assure more nearly equal availability of all turtles. These standard samples were compared by a collection ratio. Assumptions involved in the use of this ratio are discussed. Correction factors were applied to make allowance for turtles whose ranges.were partly inside and partly outside the study area, and for transient turtles. A second estimate, on the basis of the entire season's collecting, gave closely comparable results. The population of the study area is estimated to be between four and five turtles per acre, with juveniles constituting less than ten per cent of the total.