This paper presents results of research of Timberjack 240C tractor equipped with double-drum winch Adler HY16 in skidding timber asortments up and down the slope in selective felling. The work site is situated on the altidude between 870 m and 1097 m. Harvesting density was 51.86 m3/ha. Structure of growing stock was: 74.29 % of fir, 22.87 % of beech and 2.84 % of other hard broadleaves. Ž. Zečić, D. Vusić, M. Prka, S. Klepac: UTJECAJ NAGIBATRAKTORSKOG PUTANAPROIZVODNOST ...Šumarski list br. 3–4, CXXXIV (2010), 103-114 Harvesting phases of felling and processing were time-separated from skidding. Research was conducted using time and work study method. Time consumption of each work component was measured by snap-back chronometry method. Lenght of skid roads was gradually measured on uniform slope segments by GPS. Coordinates of points, where the longitudal slope changes, were recorded by GPS as well. Slope of each individual segment was calculated using the difference in altitude of slope change points and the related lenght. While skidding, the tractor was travelling exclusively by two skid roads (figure 4 and table 1). Average skid slope is expressed in percents, in the direction of loaded tractor travel. Average slope of recorded turns was calculated by weighting the related slope with the travel lenght. Average skidding distance was calculated as an average of recorded loaded trevel lenghts of each turn. Load records are shown in table 2. Significant difference in average load on skid trail 1 (5.043 m3) and on skid trail 2 (3.715 m3) is evident. By each turn difference in productivity of skidding on skid road 1 and skid road 2 increases by 1.328 m3. During the research 75.650 m3of beech wood asortments, and 115.994 m3of fir wood asortments were skidded on skid road 1. On skid road 2 18.607 m3of beech wood asortments an 44.549 m3of fir wood asortments were skidded. By multiplying achieved volumes with density of fresh beech wood (1,07 t/m3) and fresh fir wood (0,98 t/m3) average load masses on skid road 1 (5.122 t) and on skid road 2 (3.739 t) were calculated. Variable times (travel loaded and unloaded by skid roads and landing, line pulling and winching) were analysed by mathematical and statistical methods (figure 6, figure 7 and figure 8). All other effective times were regarded as fixed and calculated as averages of recorded time (figure 11). During the research average daily output of 36.40 m3/day was achieved with average total time consuption of 404,59 min/day. The said total time consumption makes 84.59 % of legal working time (480 min). With full utilisatin of legal working time, at the same organization of work, the productivity would increase by 15.41%. During the research the tractor accomplished 76.46 % of total time as efective time with delay time of 23.54 %. In effective time per turn, for skidding distance of 500 m, fixed times take 17.36 min and variable times take 18.97 min. Effective time per turn, for skidding distance of 500 m, is 36.34 min. With determined allowance time factor of 1,22 total time is 44.42 min. Average travel speed on skid roads is 3.56 km/h for unloaded tractor and 3.50 km/h for loaded tractor. Average travel speed on landing is 4.65 km/h for unloaded tractor and 4.93 km/h for loaded tractor, on average distance of 65 m. Average speed of line pulling is 0.97 km/h and average speed of winching is 0.86 km/h. Standard time of skidding on skid road 1 ranges from 5.32 min/m3(100 m) to 13.17 min/m3(1000 m) with average load of 5.04 m3. For skkiding on skid road 2 standard time ranges from 7.22 min/m3(100 m) to 17.88 min/m3(1000 m) with average load of 3.72 m3. Daily output ranges from 90.27 m3/day for skkidnig distance of 100 m to 36.45 m3/day for skidding distance of 1000 m, when skkiding on skid road 1, and from 66.50m3/day to 26.85 m3/day when skidding on skid road 2 for the same distances. For skidding distances from 100 m to 1000 m skidding cost ranges from 31.83 kn/m3 to 78.83 kn/m3, for skid road 1 and from 43.20 kn/m3to 107.01 kn/m3for skid road 2. Timberjack 240C tractor belongs to a group of highly-efficient special forest timber skidding machines. The principle of load size effect, as one of the key factors of timber skidding, on skidding productivity in different skidding distances has been proven by this detailed research. Hence, by increasing the load volume the productivity of this tractor is significantly increased, thus lowering the cost per unit, in this case by 26.34 %. Considerably lower timber skidding cost when skidding down the slope should be regarded as one of the capital elements in planning and construction of secondary forest network.