Genetic algorithms provide an extremely efficient method to assess the information on the thermal history contained in measured fission track data.

An overview of the methodology will be given, using both synthetic and real data examples. The fact that many (> 1000) thermal histories can be tested rapidly allows the spread in good data fitting solutions to be examined directly to assess the characteristics required to fit the observed data. The methodology may be applied to any type of thermochronological data, although fission track data are particularly useful because of the information on cooling rates contained in the length distribution.