Abstract Background Urban agriculture is potentially an important piece of the food security puzzle for a rapidly growing urban world population. Community gardening is also promoted as a safe and viable form of exercise for aging populations in crowded settings where opportunities to participate in other action activities may be limited. Knowledge of potential site-specific health risks to environmental contaminants is important in dialogues promoting urban farming. Methods We assess the pseudo-total concentrations of selected potentially toxic elements (PTEs) in the soils of community gardens, public parks, and woodlands in the tropical urban island nation of Singapore. We compare concentrations of cadmium, copper, lead, and zinc with amalgamated risk guidelines to form a baseline understanding of the level of contamination in these spaces. We also perform providence tracking with lead isotopes to identify potential sources of contaminants. Results All pseudo-total concentrations of Cd, Cu, Pb, and Zn in the soil were below threshold concentrations considered to represent substantial risk. Further, PTE concentrations in gardens were largely equivalent to those found in community parks and woodlands, but the geographical distribution varied. Provenance tracking with Pb isotopes indicated Pb in gardens was both anthropogenic and natural, but spatially variable. The lack of strong spatial clustering of areas with the highest PTE concentrations was inconsistent with a common point source of contamination. However, the correlation between Cu and Zn suggest a common source for these elements, such as road/trafficking or atmospheric deposition. Conclusion We find limited risk of urban gardeners to exposure to Cd, Cu, Pb, and Zn—elements that are commonly abundant in urban settings with dense transportation networks and substantial industrial activities. The low levels of PTEs are encouraging for the promotion of urban farming for food production and leisure in this dense urban setting. However, as concentrations were low, we did not assess bioavailability and bioaccessibility of the PTEs. These assessments would need to be determined in cases of with higher levels of contamination to provide a more thorough consideration of actual human risk.