Humans are inherently drawn to places of naural scenic beauty. Alone in Australia's national parks; the demand on these locations is increasing because of increased population plus increasing regional domestic travel. It is important to understand how tourists are utilising nature-based tourism resources to ensure that tourist attractions, facilities and services provided in parks are not exposed to loss events. Environmental and human losses, when combined with their likelihood of occurrence, will constitute risk. Heavy use of popular sites in particular times of the year, beyond the carrying capacity, may result in an irreparable environmental loss. Besides, human risks in terms of becoming lost in the wilderness or falling down from the cliff are significant factors that park managers are considering. It is becoming utmost important to investigate how these visitors are interacting with the natural environment, in what ways at highly vunerable sites. Park managers are now required to understand the needs, demands, preferences and motivation of visitors to vist the parks. A proactive planning to manage potential risk particularly in relatively more fragile and vulnerable environments is needed. This thesis documents a study of visitors to Loch Ard Gorge Site within the Port Campbell National Park, Victoria. Visitors were monitored for their spatial behaviour using the GPS receivers. Visitor's attitude, interests, opinions and motives to travel to the study location is surveyed using onsite self-administrated questionnaire. Total 102 individuals were surveyed to determine their preferences and to obtain their socio-demographical profiles.These were latter coupled with their GPS track movements in order to derive their elicited spatial behaviour. It is well-established fact that visitors' attitude, preferences, and socio-demographical profiles are important factors to identify risk taking spatial behaviour. Hence, this research underpins the exhibited spatial behaviour of tourist to deliver their associated spatial behavioural risk typologies. Statistical classifications of visitors, based on survey responses were categorised using K-means Cluster Analysis. As a result, six different tourist types were obtained. They are; • Mid-Allocentics possibly Risk Takers. • Allocentrics and confirmed Risk Takers. • Mid-Psychocentrics and Risk Averters. • Allocentrics and Dependent Risk Takers. • Psychocentrics and confirming Risk Averters. • Psychocentrics and Risk Averters Mass Tourists. The mean values of all clusters (cluster centroids) are interpreted to understand tourist's risk taking and/or risk averting preferences. Visitors' actual spatial behavioural were spatially analysed using movement pattern maps. The typologies subsequently were verified using Discriminant Function Analysis and in this process tests of equality of group means (Significance of F-test), ANOVA classification are discussed. The final part of Discriminant Function Analysis is to determine the linear regression equations for prediction of group membership of data points in future using classification Function Matrix (Fisher's Linear Disceriminant Function). The developed tourist typologies may help park managers to regulate and mitigate human risk prior to its occurrence by understanding the visitors personality and preferences and their risk taking probabilities.