Practical experience has indicated that shooting water buffalo with a captive bolt gun in the front of the head does not always produce an effective stun. Slaughtermen have been claiming that the poll position is more reliable, but under present EU regulations this shooting position is not allowed for domesticated bovines. This study examined the effectiveness of shooting water buffalo with captive bolt guns in the poll region. The depth of concussion was assessed in 30 water buffalo from physical collapse, presence or absence of corneal and palpebral reflexes, normal rhythmic breathing, eyeball rotation, nystagmus and whether the animal was re-shot. All except one animal collapsed immediately. The one animal that failed to collapse was not breathing. The prevalence of a shallow depth of concussion in the poll shot animals was 53%. Two had been shot through the spinal cord, and with the exception of those plus one other in which the bolt did not penetrate the brain, the other animals were considered to be adequately stunned. Post mortem examination of the brain showed that 79% of the under thirty month old animals were shot in the midbrain or brain stem, but these regions had to be avoided in over thirty month old animals because of the risk of damaging the brain stem sample needed for BSE testing. In addition, three water buffalo were shot in the crown position and one in the front of the head. The frontal position was ineffective as the animal did not collapse, and all animals shot in the crown position resumed breathing shortly after shooting, indicating a shallow depth of concussion. It is concluded that poll shooting can be effective in water buffalo, but it produces a shallow depth of concussion compared with frontal shooting in cattle. It requires accurate placement of the gun to ensure that buffalo are not shot through the spinal cord instead of the brain. Vigilance is needed in ensuring that the animals are stuck promptly so that none recover consciousness.