There is an established consensus among researchers that contact with nature improves mental health, wellbeing, and quality of life in urbanised environments. Studies tend to examine the health impacts of nature without identifying specific physical and spatial landscape features that could guide health-promoting design of urban green spaces. A growing body of evidence suggests that landscape features described in the Contemplative Landscape Model (CLM) can be used to measure therapeutic value of urban landscapes. CLM assesses urban landscapes across seven sub-scales: Layers of the Landscape, Landform, Vegetation, Color and Light, Compatibility, Archetypal Elements and Character of Peace and Silence. We exposed 74 healthy adults to six urban landscapes in laboratory (video representations) and naturalistic outdoor settings. We explored the associations between the visual quality of urban landscapes annotated with CLM, with self-reported positive emotions and brain activity consistent with mindfulness (Theta waves), relaxation (Alpha waves) and attention restoration (Beta waves), and differences between laboratory and naturalistic setting. CLM scores predicted self-reported Valence and Arousal, and low frequency power bands: Alpha and Theta in the naturalistic setting. Landscape features showing the strongest associations were Character of Peace and Silence, Layers of the Landscape and Archetypal Elements. Alpha, Theta brain reactivity and Arousal scores, were significantly different between laboratory and naturalistic settings (p < 0.05), while Valence scores between those settings were statistically identical (p = 0.22). Self-reported Valence and Arousal, but not brain activity, were significantly associated with the majority of landscape features in the laboratory setting. The results of the study provide guidelines on the urban landscape features most beneficial for human health, to inform urban green space design.