While it has been established previously that the strengths of commercial timber products containing knots and other natural growth characteristics are greatly influenced by the physical dimensions of the pieces, it has not been established whether the same is the case for defect-free wood. The structural engineer has a choice of many grades of timber available for his designs and it may be argued that these grades form a stepwise progression from the rather weak No. 3 grade through select structural grade and into clear defect-free sawn wood. For efficient design, it is important to have the size effects correctly quantified for the different qualities. This study concentrates on quantifying the size effect at the strong end of the perceived progression of increased strength, the clear defect-free wood. For the bending tests carried out in this study, it was found that the length effect was very pronounced while thickness did not affect strength. A consistent depth effect could not be found. If some depth effect did exist, it would be of second order in importance relative to the length effect. From the tests, it was established that the length effect parameter is unlikely to be less than 0.17 for bending. The work was conducted using the species group spruce–pine–fir, which is the most abundant species group in Canada. The tests add further evidence to the argument that the effect of length must be included in the design process for any material where the strength varies along the length of the member. Key words: size effects, defect-free wood, spruce–pine–fir, adjustment factor, length effect parameter.