The main scope of the International Physical Activity and the Environment (IPEN) study is to examine the strength and shape of relationships between environmental characteristics and physical activity using pooled data from 17 cities representing 12 countries across 5 continents. Pooled analyses of data from the IPEN study are considerably challenging because they need to (1) account for multiple sources of dependency in the data; (2) model non-normally distributed outcomes; (3) allow for non-linear relationships; (4) allow for differences in associations at the within- and between-site levels; (5) and deal with missing data. This interactive workshop identified and described challenges encountered in conducting pooled analyses of the IPEN study and presented analytical approaches that can be applied to any study facing similar complex problems. Participants had the opportunity to run various generalized additive mixed models in R appropriate for correlated, non-normally distributed and non-linearly related data using their own data or data provided by the presenter.