In this study five parasitological methods and a polymerase chain reaction (PCR) were compared for the diagnostic sensitivity for Trypanosoma evansi in experimentally infected water buffaloes over a period of 15 weeks. The combined estimates of sensitivity (CE(se)) of the PCR proved to be highest at 78.2%, closely followed by the mouse inoculation (MI), the micro-haematocrite centrifugation technique (MHCT) and the mini-anion-exchange centrifugation technique (MAECT) with CE(se) of, respectively, 74.0, 69.6 and 62.4%. The CE(se) of the buffy-coat technique (BCT) at 38.6% and the sodium dodecyl sulfate (SDS) clarification technique at 25.1% were considerably lower. PCR detected consistently all buffaloes infected from week 3 post-infection (PI) onwards. For MI this occurred after 5 weeks PI while for MHCT and MAECT these sustainable high levels were reached in the 7th week PI. BCT and SDS never detected all buffaloes infected. The influence of time and temperature on the viability of T. evansi in heparinized blood from water buffalo was also studied. In general we observed that the survival time tends to be longer when blood is kept at 4 degrees C. In samples kept in direct sunlight parasites became undetectable with the MHCT after 30min. After treatment of the water buffaloes with diminazene aceturate, the PCR signal disappeared within 24h.