The present study was conducted to assess the interrelation between teat anatomy and machine milking in dairy buffaloes raised in Switzerland. A 3-min pre-stimulation induced milk ejection before cluster attachment in most cases and caused an optimal milk removal during machine milking. In an additional experiment, longitudinal cross-section ultrasound was obtained before and after a 3-min pre-stimulation. Teat wall thickness, teat diameter, cisternal diameter and teat canal length were evaluated. It was observed that 3-min pre-stimulation dramatically reduced teat canal length whereas all the other anatomical parameters remained unchanged. The vacuum needed to open the teat canal was also measured before and after a 3-min pre-stimulation by using a special teat cup with only the mouthpiece of the liner remaining on the top of the teat cup (no liner, no pulsation). Without pre-stimulation but after wetting the teat canal by stripping one squirt of milk out of the teat, no milk could be withdrawn with a vacuum up to 39 kPa. However, after pre-stimulation, milk flow occurred in all buffaloes at a vacuum between 16 and 38 kPa. In the last experiment, the teat tissue was examined in slaughtered buffaloes and compared with teat tissue of cows. No difference was noted in histological sections and teat canal length was similar in cows and buffaloes. Proximal to the teat canal, the teat did not pass into an open cistern but the lumen was collapsed. In conclusion, buffaloes need to be well pre-stimulated because the tissue above the teat canal provides additional teat closure before milk ejection. Therefore, milk can only be obtained after pre-stimulation.