Exposure to green space has been suggested to reduce the risk of several adverse health outcomes by multiple pathways including physical activity (James et al., 2015). Green space could reduce the likelihood of being overweight by offering suitable spaces that encourage physical activity (Hartig et al., 2014; Lee and Maheswaran, 2011). Overweight and physical inactivity are important risk factors for cardiovascular disease, diabetes and mental illness (Lachowycz and Jones, 2011). Green space is inconsistently associated with physical activity and being overweight, according to reviews of epidemiological studies (Hartig et al., 2014; James et al., 2015; Lachowycz and Jones, 2011; Bancroft et al., 2015; Nieuwenhuijsen et al., 2017; Durand et al., 2011; Hansen et al., 2015; Mayne et al., 2015; Hunter et al., 2015). Green space was negatively associated with being overweight in some studies (Liu et al., 2007; Pereira et al., 2013), however in other studies, green space was either not associated or positively associated with being overweight (Mowafi et al., 2012; Richardson et al., 2013; Picavet et al., 2016; Cummins and Fagg, 2012; Witten et al., 2008). These inconsistent associations across studies have been attributed to the use of different green space measures, the focus on quantity instead of quality and the use of green space, different populations and definitions of physical activity (e.g. self-reported vs. accelerometry, total versus vs. outdoor) and lack of control for important confounders (Lachowycz and Jones, 2011; James et al., 2015). Studies in large populations evaluating some of these sources of heterogeneity are needed. The most commonly used way to objectively assess exposure to green space is to determine the greenness within a circular buffer of the residential address (James et al., 2015). The radius of the buffers used to define surrounding green space varied among studies and ranged from 30 (McMorris et al., 2015) to 3000 m (Maas et al., 2008). Green space within a buffer can be assessed by using the Normalized Difference Vegetation Index (NDVI) or by using national land-use databases (James et al., 2015). Another common measure to determine exposure to green space is the distance (Euclidian or network) of the residential address to the nearest park entrance (James et al., 2015). So far, only a few epidemiological studies have investigated the effects of different types of green space (Hartig et al., 2014; Lee and Maheswaran, 2011). Green spaces can be subdivided into publically accessible and private green; or into natural, agricultural and urban green. Variance in greenness within the surrounding area can also be used as a measure (Pereira et al., 2013). In addition, most studies assumed a linear association of exposure to green spaces with the health outcome of interest (Shanahan et al., 2015) without providing information about the validity of this assumption. Non-linear associations between green space and overweight and physical activity are plausible. For instance, high levels of green space tend to be associated with higher distances to facilities (e.g. shops, work). Distance is probably the most important determinant for choosing an active or non-active mode of transportation (Heinen et al., 2010). Assessing exposure-response curves could provide information on how changes in quantity of green space could affect overweight and physical activity (Shanahan et al., 2015). The goal of this cross-sectional study is to evaluate whether the association between exposure to green space and being overweight (body mass index (BMI) ≥ ۲۵ kg/m2 ) and outdoor physical activity depends on the definition of green space, adjustment for potential confounders, population characteristics and assumed shape of the association.