A method for pollen-based geolocation. The method determines the probability P that a given location is part of the travel history of a given sample. Using simulated datasets and Monte Carlo simulation, the model parameters can be precisely associated with P, thereby allowing the algorithm to operate on real-life samples of interest.