The literature on restorative environments has analysed the positive effects of people’s transactions with different natural (wild areas, urban parks, etc.) and built settings (museums, monasteries, historical towns, etc.). To date, little attention has been paid to the analysis of the restorative potential of botanical gardens, which are characterised by a high level of biodiversity and human-made elements at the same time. Recent findings support the idea that biodiversity can play a major role in users’ evaluations of green areas (Chen, Adimo, & Bao, 2009; Lindemann-Matthies & Bose, 2007) and may increase the psychological benefits associated with experience in nature (Fuller, Irvine, Devine- Wright, Warren, & Gaston, 2007). In addition, little is known about the process leading to restoration, as studies have mostly addressed the analysis of positive outcomes of restorative experiences. Botanical gardens with distinct physical properties were identified in four Italian cities, namely Padua, Florence, Rome, and Bari, and considered for the study. We administered a questionnaire focusing on people’s experience in the botanical garden (length and frequency of visits, activities performed, perceived restorative properties and restorativeness, emotional response, perceived well-being) to an opportunistic sample (N = 127) contacted at the different locations. ANOVAs were conducted to analyse the relationships between the physical properties of the four botanical gardens and people’s on-site experience, in terms of perceived restorative properties, emotional response, perceived well-being. In addition, we analysed the effect of group characteristics (gender, age, working activity) and of activities performed on users’ experience in the botanical garden. A multiple regression analysis predicting perceived well-being was then carried out to test a model including several sets of variables: structural (e.g. distance from home, working activity, etc.), experiential (length and frequency of visits, activities performed), and psychological (perceived restorative properties, emotional response) ones. Results showed that the physical properties of botanical gardens affect the perception of the restorative properties and the emotional response toward the environment. Also activities performed play a role in shaping the experience of users: the more users spend their time in the botanical garden, the higher is the perceived level of the restorative properties and the better is the emotional response. Finally, multiple regression analysis showed that well-being experienced in botanical gardens is predicted by variables from the different sets included in the model, with perceived restorativeness and emotional response playing a key role. In particular, our analysis outline a mediation role of perceived restorativeness and emotional response in the relationship between experiential variables and perceived well-being. Theoretical implications for the study of restorative environments, and suggestions for the management of botanical gardens are discussed.