Tool wear is a major issue in the forest industry, and extensive research has been carried out to understand its causes and to limit tool degradation. Among these causes, the properties of the machined wood — highly variable between species — play a crucial role. Sixty years ago, André Chardin (d. 1987) conducted a 15-year experimental study on tool wear during the sawing of tropical woods. Out of 4,000 cutting tests, 835 standard tests with uniform chip thickness were performed on 554 trees representing 285 species, 200 genera and 59 families, at both green and air-dry moisture contents. The tested woods were highly diverse, reflecting the wide variability typical of tropical species. Tool wear varied greatly, partly due to plastic bending of the cutting edge under high mechanical stress. The relationship between tool tip recession and cutting length followed a power law with a median exponent of about 0.7, which was lower for tests showing tip bending and for dry wood compared with green wood. Silica content emerged as the dominant factor influencing wear, given its large variation among species, while extractive content was not significant. Wood density and moisture content also affected wear, though to a much lesser extent. Grouping the tests by moisture (dry versus green) and silica content (siliceous versus non-siliceous) highlighted the combined effects of density and moisture on edge recession. Tool tip bending occurred frequently (42%) for dry non-siliceous woods, less often (10%) for dry siliceous or green non-siliceous woods, and rarely (3%) for green siliceous woods. Tool face burning was also common (42%) in dry non-siliceous woods. For the 78% of tests without tool tip bending, progressive linear regression on logarithmic wear and wood parameters yielded a 58% predictive value: silica (42%), moisture (8%), specific gravity (7%), and extractives (0.4%). The minimal impact of extractives contrasts with their known influence on steel corrosion and frictio