Measured wear data is presented for 3 different carbide grades. The data were collected during rip sawing wood with a double arbour saw. The purpose of the test was to determine the suitability of different grades for sawing frozen timber. A set of circular saw blades of diameter 350 mm was equipped with teeth comprised of 3 different cemented carbide grades, denoted A, B and C. These 3 grades were chosen out of 6 grades in an earlier preliminary laboratory test. Grade A was a relatively soft and tough standard grade (K1C=17, HV30=1150) normally used in sawmills in winter conditions for frozen wood, grade B was harder and more brittle (K1C=11, HV30=1600) and grade C was even harder and more brittle (K1C=9.5, HV30=1950). The double arbour saw (with vertical arbours) was equipped with 6 saw blades (3 on each arbour) for cutting 2 centre boards and 2 side boards. The 6 saw blades with different teeth were mounted in a mixed manner on the arbours, and after sawing a number of logs the wear of teeth was measured. After some time the blades were removed, inspected, ground and used again for sawing. The thickness of boards was also measured and the standard deviation was calculated.The results show cutting edge radii as a function of the number of logs sawn and also the standard deviation of the thickness of the sawn boards. Grade A had the highest wear and grades B and C the lowest wear. There was no significant edge damage during the tests. Grade C did not suffer problems of chipping from cutting edges and was found to be suitable for sawing frozen timber. The thickness standard deviations were constant at about 0.2 mm, and not a function of the number of logs sawn.