Noninvasive samples have proved useful in genotyping studies of free-ranging mammals. However, potential genotyping errors associated with such samples dictate the need for validation studies. This pilot study demonstrates the use of dolphin faeces in multilocus microsatellite genotyping studies. An empirical approach to calculating the rate of genotyping error was applied to data from matched pairs of blood or tissue and faecal samples from both captive and wild bottlenose dolphins. Microsatellite genotypes were assigned to dolphin faecal extracts with greater than 95% confidence by using a multiple tube approach, and at least two independent replicate genotypings.