Gambling is an important recreational activity in New Zealand, with high levels of participation by the general public. Although gambling activities are an important source of employment and a means of raising funds for various community and sporting purposes, gambling on electronic gambling machines (EGMs), both in casino and non-casino venues are known to be correlated with gambling-related harm, resulting in higher levels of personal, familial, health and societal problems. After undertaking a review of relevant literature on participation in gambling activities and accessibility to gambling venues, it was found that although studies examining the accessibility of venues with EGMs have been researched to some extent in an overseas context, studies pertaining to such venues in New Zealand have been limited. This study therefore aimed to investigate the link between accessibility to gambling venues with EGMs, including distance-wise proximity to such venues and the number of these venues within a certain distance, and their impact on gambling behaviour of individuals. Two measures of accessibility were investigated; the distance from a participant’s residential location to the nearest casino and non-casino EGM venue, as well as the number of venues with EGMs within 500 metres, one kilometre, five kilometres and 10 kilometres, taking into account the travel network including roads, walkways and ferry routes. Three outcome variables were considered, in particular past-year gambling, gambling on EGMs at either a casino or non-casino venue and the risk of problem gambling for past-year gamblers. Logistic regression analyses were carried out on each outcome variable to test the significance of each distance and density variables separately after correcting for individuals’ sex, age group, prioritised ethnicity, neighbourhood deprivation and individual deprivation. A final multiple variable model was then created for each of the three outcomes using the best subset model building process. Final multiple variable models for all three outcomes found distance to the nearest casino to be an important predictor, signifying the importance of casinos as destinations for entertainment and tourism, while those predicting past-year gambling on EGMs and the risk of problem gambling additionally found the number of venues within 10 kilometres to be a strong predictor, thus signifying the importance of convenient driving distance in reaching their desired venue. The results of the multiple variable models were successfully illustrated using maps of different geographical areas in New Zealand.