The purpose of this study was to clarify whether the wood grinding model – based on an energy balance for the grinding zone – would improve understanding of wood grinding for pulp. This study relied on previously obtained data by the Finnish Pulp and Paper Research Institute. The frictional coefficient (Pc/Pt) computed and the power-specific groundwood production (Ġw/Pt) were important x- and y- variables, respectively. Fresh spruce wood samples were ground by application of a laboratory grinder, where the stone surface speeds were 30, 15 and 7 m/s, respectively. The power-specific productivities of high- and medium-speed grindings followed one and the same mechanism, since both speeds led to a productivity of 0.99 [(kg/h)/kW]; the low-speed grinding, however, led to a level of 0.66 [(kg/h)/kW] at frictional coefficients closest to 100 mW/kW. The rate of formation of coarse rejects was – at the same frictional coefficient as before – 15.0, 4.3 and 2.3 mg/s for high-, medium- and low-speed grindings, respectively. However, the rate of fines formation determined by McNett apparatus was about ten times higher than that of formation of coarse rejects: 123.0, 42.2 and 23.7 mg/s, respectively. The fines-to-shives ratio (determined by a Somerville shive analyser) was assumed to indicate fiberisation for high-, medium- and low-speed grindings, and the true data, most close to 100 mW/kW, were 56.0, 55.9 and 36.1 units of fines-to-shives, respectively. The curves followed the same trend, but on slightly different levels. As for the important sheet properties, the tensile strengths of high-, medium- and low-speed grindings were low, medium and high, respectively: 37.1, 46.9 and 56.2 Nm/g. The light scattering coefficients of high-, medium- and low-speed grindings were low, medium and high, respectively, or as data being most close to 100 mW/kW: 59.2, 66.0 and 67.5 m2/kg, respectively. Some general conclusions may be drawn from these results. To achieve the best groundwood productivity, the frictional coefficient should be kept on a level close to 100 mW/kW. Generally speaking, it seems that high- and medium-speed grindings appeared to act as following the same mechanism as far as the productivity was concerned, but the rates of shives and fines formation did not follow such a pattern. The groundwood sheets showed lower tensile strength in high-speed grinding than in medium-speed grinding, while the light scattering coefficient was much lower for the high-speed grinding than for the medium-speed grinding. Because of its low productivity, the low-speed grinding does not seem to be useful, although high tensile strength and high light scattering of the sheets would plead for it.