) trees from one fast-grown and one slow-grown stand in southern Sweden. From the trees 240 studs (45 × 70 × 2500 mm) were taken for measurement of distortion. Wood properties were measured on small specimens (13 × 13 × 200 mm) cut from the studs. Spiral grain angle was found to vary from approximately +3° (left-handed) close to pith to zero 150 mm from pith with a strong individual variation. The material from the fast-grown stand had a larger spiral grain angle compared with the slow-grown material. Spiral grain was poorly correlated to other parameters. Presence of knots had a substantial influence on longitudinal shrinkage (αl) measurements. Specimens with large knots (KAR > 33%) had almost 100% higher longitudinal shrinkage than specimens without knots. It should be pointed out, however, that measuring shrinkage in small specimens containing even small knots can create a problem with regards to the obtained results, especially results of αl. It was found that presence of compression wood in several growth rings more than doubled the longitudinal shrinkage. For the radial and tangential direction the presence of compression wood decreased shrinkage with about 30%. The ratio between tangential and longitudinal shrinkage was 49 for normal wood whereas for compression wood the ratio was 13. These results confirm the theory that the microfibril angle governs shrinkage. Longitudinal shrinkage decreased slightly with increased distance from pith whereas radial and tangential shrinkage did not display any substantial radial variation. The fast-grown material had generally a higher longitudinal shrinkage and lower transverse shrinkage than the material from the slow-grown stand. About 50% of the variation in longitudinal shrinkage was explained by radial position, density and ring width. Density and ring width did explain 60% of the variation in radial shrinkage but only 30% of the variation in tangential shrinkage.