Interpretation is widely used as a management tool because: it is perceived to be the most cost effective method; it is a light-handed approach and allows visitors the freedom of choice; and it enhances visitor experiences and satisfaction. However, the ability of interpretation to bring about a reduction in the impacts of visitors to natural areas, has rarely been quantified. This study was designed to determine the extent to which an interpretive program reduced the environmental impacts of visitors to national parks. Fieldwork was undertaken in Lamington National Park, where 41 guided walks were examined. To measure the actual behaviour or resulting impacts of visitors in a national park, three appropriate environmental indicators were chosen: shortcutting of corners, picking up litter already on the track, and noise levels. Five interpretive programs were created, each with a different combination of environmental interpretation, role modelling and verbal appeals. For the shortcutting results, the interpretive program with the combination of environmental interpretation, role modelling by the guide and verbal appeals from the guide, was always the most effective in reducing shortcutting. Visitors in this interpretive program were always, statistically, less likely to shortcut than visitors on all the other interpretive programs. The programs with only environmental interpretation or no environmental interpretation, were always least effective in reducing shortcutting. The interpretive programs with environment interpretation plus role modelling, or verbal appeals, were always in the middle of these extremes. They were more effective than having neither role modelling or verbal appeals, but less effective than having both. Results for the amount of litter picked up found that the inclusion of verbal appeals in an interpretive program was the only factor that influenced whether visitors picked up litter. In addition, there was no difference in the amount of litter picked up, by the interpretive program with environmental interpretation only, and the program with no environmental interpretation. Results for the noise levels of visitors, found that no interpretive program reduced the amount of shouting and talking loudly of visitors. Although not statistically significant, it did appear that there were lower proportions of shouting and talking loudly, following a verbal appeal and/or role modelling. Additionally, there was no influence of the interpretive program on the proportion of time visitors were talking, or quiet, during their walk. Overall, this research found that interpretation can be an effective management tool in reducing visitor impacts. Interpretation is most effective in reducing impacts when those impacts are specifically addressed through verbal appeals, combined with positive role modelling of appropriate behaviours. However, interpretation did not reduce all the impacts studied and therefore is not the solution to all problems. Implications of this study are that those using interpretation as a means of reducing visitor impacts, must ensure that they have a high standard of interpretation, which specifically addresses the impacts that need to be reduced. It also highlights the importance of the role of the guide, and that those employed should be well trained and competent in their position.