Individual variability in the flight potential of Ips sexdentatus Boern. was investigated using flight mill experiments. Non-flyer status and flight durations were studied in relation to day of emergence. sex, pronotum width, dried body weight (after lipid ether extraction). and lipid content estimate. Lipid content estimation was determined in beetles reared on the same trees and under identical conditions as the flight test insects. The four flight-tested cohorts showed that approximately one-third were non-flyers. The distribution of individual flight durations was described by a log-normal curve. Consequently. the dispersal variability within the population could be fitted to a linear model on a log-probit scale. The concept of FD50 (flight duration 50). defined as the flight duration of 50% of a sample population. was used to describe population flight potential. There was no significant correlation between the individual flight duration and the day of emergence. sex. body size or weight, or lipid content estimate. A significant positive correlation was recorded between the different classes of flyer (i.e. non-, short, long, and very long flyers) and the average lipid content estimate. It was hypothesized that (1) a mininium threshold fuel supply was necessary to initiate the dispersal flight and (2) fuel supply could result in dispersal tendencies representing a specific response to habitat constraints.