In three controlled-environment experiments, diffuse knapweed plants were vernalized at leaf numbers ranging from 1 to 20. Five weeks after return to 16 h light: 8 h dark cycle at 25:20 °C (light:dark), bolting (elongation of flower stalks) was recorded. In a separate field experiment we marked 200 rosettes with leaf numbers ranging from 4 to 26 in November, and on June 30 we recorded which of these had bolted. Probit analysis was used to derive an LN50 (leaf number at which 50% bolting is expected) for each trial. Vernalizing conditions strongly affected the LN50 values derived, mainly through their influence on leaf development. The lowest value of LN50 (5.9 leaves) was obtained with vernalization at 7 °C and 460 μmol m−2 s−1 light intensity, an intermediate value (LN50 = 8.9 leaves) with vernalization at 8 °C and 164 μmol m−2 s−1 intensity, and the highest value (LN50 = 12.6 leaves) with vernalization at 4 °C and 164 μmol m−2 s−1. These experimental conditions allowed an average of 6.5, 4.2, and 1.7 leaves to form during vernalization, respectively. Presumably the vernalizing treatment that allowed the least leaf formation (lowest temperature and light intensity) gave the most reliable estimate of the end of the juvenile period, ending at the formation of the 13th leaf (LN50 = 12.6). Key words: knapweed, vernalization, juvenility, leaf number, bolting.