Stoats (Mustela erminea, Linnaeus 1758) are a major conservation pest in New Zealand, and they are presently the focus of much research to improve the efficacy of their control (Department of Conservation, 2000). Tunnels are used in almost all aspects of monitoring and control associated with stoats, yet few studies have investigated the influence of tunnel design on stoats' behavioural responses to them, or the effect on capture rate (Dilks et al.,1996; Maxwell et al., 1997; Short and Reynolds, 2000). This thesis describes pen and field research, which examines the influence of several tunnel design variables on stoat entry behaviour. This information will be used to assist in designing a trapping tunnel for a new repeat-kill, permanent-set kill trap for sustained control of stoats in indigenous forest, currently being developed at Lincoln University. The investigative behaviours of captive stoats were videoed in large outdoor pens. I investigated the effects of diameter (50, 100 and 150 mm), length (400, 600 and 800 mm) and-end type (open, closed and mesh) on stoat entry behaviour into a variety of PVC pipe tunnel types. Three hair collection methods, suitable for monitoring stoats, were also trialled both in the pens and in the field. Initial and repeat entry behaviour was observed as well as general investigative behaviour toward the tunnels. Diameter does affect repeat entry type and frequency in closed ended tunnels, however, the comparison of tunnel diameters in open-ended tunnels demonstrated that diameter alone may have a minor effect on stoat entry behaviour. Diameter did not affect initial entry behaviour into any of the tunnel types and had no influence on the depth to which stoats entered the tunnels. Longer tunnels may encourage more body entries into smaller (50 mm) diameter tunnels. However, stoats were reluctant to proceed to the end of the tunnels of any length. End type affects both initial and repeat entry behaviour in all tunnel types, although it is more apparent in tunnels with a smaller diameter. Removal of the end cap resulted in significant changes in repeat entry behaviour into the 50 and 100 mm tunnels. Addition of a mesh end cap appeared to discourage stoats from entering 50 mm tunnels. During field trials, stoats were detected in both open and closed-ended tunnels, but more frequently in closed-ended tunnels. The hair collection methods successfully collected adequate hair samples in the pen trials but did not perform as well in the field. Tunnel design does influence stoat entry behaviour and all new tunnel designs should undergo testing prior to use in the field. With no difference in initial entry behaviour, a small (50mm) closed-ended tunnel is likely to reduce non-target entries and position the stoat correctly for a humane kill with the new kill device.