Based on the optimal partitioning theory, the comparative assessment of seedling allometry is a common task in retrospective genetic tests and early testing of forest reproductive material. Our hypothesis was that root restriction imposed by the container might hinder or rule out genetic differences in biomass allocation. We grew seedlings of two contrasted provenances of Canary Islands pine in mini-rhizotrons, 60 and 90 cm deep, and alternatively in standard bottom-open 200 cc forest containers. In the mini-rhizotrons, plants from the drier provenance allocated more biomass to roots, especially to the tap root and invested less in needles, both in a biomass and leaf area basis, and this morphological divergence increased between two harvests, undertaken at 57 and 115 days after planting. By contrast, confirming our hypothesis, at the 115 days harvest, the plants grown in standard containers did not exhibited significant differences between provenances for Leaf Mass Fraction, Root Mass Fraction and Leaf Area Ratio. We conclude that the physical constraint for root development imposed by small containers increases the probability of dismissing the genetic effect in biomass allocation when assessing forest reproductive materials at the short term, even when the whole plant growth (total dry weight) might be unaffected.