Two visitor-monitoring methods, video monitoring and counts by human observers, were compared in order to identify and evaluate their respective advantages and disadvantages. The analysis considered user numbers, user type (walkers, dog walkers, bicyclists, and joggers) and group size of visitors. Remarkable differences were found between the two methods for user type and use levels. At low use levels, evaluations based on video monitoring resulted in fewer single bicyclists compared to counts by human observers, whereas at high use levels, human observers counted fewer walkers and bikers than video-interpreters. Based on this comparative analysis, we derive recommendations for more effective visitor-monitoring approaches. All data were collected during a visitor-monitoring project in the Danube Floodplains National Park in Austria between 1998 and 1999.