The aim of this work was to examine the effects of sampling method on accuracy of culture for Tritrichomonas foetus; and the effects of sampling method, time delay, and addition of a DNA preservative on the accuracy of a PCR test. Samples from two different sources were used for Experiment 1. Preputial scrapings were collected from one group of three infected and one uninfected bull 10 times. Secondly, samples were collected from 5 infected bulls by both sheath washing and scraping on 6 occasions, while 8 uninfected animals were sampled three or more times. Twenty nine out of 30 samples from the first sample set were found to be positive, and 83 % of samples collected by both methods for the second sample set tested positive. No samples from the control bulls were found to be positive. Scraping was found to offer significant practical advantages over washing. It may be subject to greater operator variability than sheath washing. The second experiment utilised the same samples as were used for the second data set under Experiment 1. Guanidinium thiocyanate (GuSCN) was added to half of each sample. Each sample was cultured, while all samples were subjected to DNA extraction within 6 and 30 hours and after 5 days of storage at 4 °C. PCR and agarose gel electrophoresis was performed. No samples from the control animals tested positive on PCR. The sensitivity of the PCR on samples from infected bulls ranged from 0,9 in samples extracted within 6 hours to 0,31 in samples extracted after 5 days. Sampling method had no effect with the exception of samples held for 5 days with GuSCN, where sheath washing was superior to scraping. The addition of GuSCN had no effect. Holding time reduced sensitivity at 5 days, but the effect was not significant at 30 hours. It is concluded that preputial scraping is equal in sensitivity to washing for culture of Tritrichomonas foetus. Preputial samples for PCR testing should be submitted as soon as possible after collection, and the addition of GuSCN has no effect. Samples collected by sheath washing may be superior to those collected by scraping for PCR testing. The requirement for a test with sufficient sensitivity to allow reliable identification of infected bulls based on one sample has not been met with the described method.