Viral and bacterial genomes accumulate mutations over short time spans, and pathogen sequences can be used to create time scaled trees in order to track the spread of infection. Classifying pathogen sequences from individual hosts into location (or host-type) demes, and including the deme labels as discrete traits upon phylogenetic trees allows the transition rates between demes to be estimated; which can be interpreted as a transmission network. Alternatively, locations may be included as continuous traits, diffusion rates estimated and spatial transmission routes inferred. However, these time resolved bayesian phylogeographic methods whilst excellent for small well sampled data sets are too computationally demanding for data sets of more than a few hundred sequences, and biased sampling leads to biased results. Therefore I will discuss how inferred transmission patterns from a simpler faster methodology based on repeated subsampling of sequence datasets compares to time resolved bayesian phylogeography, and show results from bovine viral diarrhoea, foot-and-mouth disease and avian influenza studies.