The biomass ofPinus sylvestris, Pinus contortaandPinus cembraseedlings that were raised in three different growing media, as well as of naturally regeneratedP. sylvestrissaplings, was estimated using functions based on variables measured non‐destructively. These included stem diameter, seedling height, number of needles and branches, length of the longest needle, and projected seedling or sapling crown area. Projected area of a two‐dimensional video image of the seedling or sapling was measured using computer‐based image analysis. The variables were tested in all possible combinations to achieve the best biomass estimate. Projected area alone explained more than 97% of the variation in seedling or sapling mass. Addition of stem diameter increased the degree of explanation to more than 98%, whereas the other variables did not contribute further. Separate functions for species and growing media resulted in more precise estimates than if general functions were used.