The use of low-grade logs to build spirally wound laminated veneer lumber (LVL) has been studied and improved from the point of view of the gluing process, fiber orientation angle, and end joint of the LVL. The butt joint appears to be the fracture point when the column is submitted to a compressive or bending load. Owing to the complexity of cylindrical LVL, we used a finite element method to simulate the mechanical behavior of part of its wall. This part was small enough to be considered flat but was representative of the structure, especially in the area of the butt joint. This allowed us to test the validity of different settings of the parameters involved in the manufacturing process. To feed data for this model, we used the results established for the linear and nonlinear behavior of raw hinoki in Part I of this series of articles. We then used this numerical model to improve the quality of the butt joint by testing different settings of the joint. We show that reducing the butt joint gap under 0.5 mm, which requires only a few changes in the production line, provides an important increase in the modulus of upture and nonnegligible improvement of the modulus of elasticity compared to that for a ≥ 1 mm butt joint gap.