Despite the growing literature on loneliness, little attention has been paid to the impact of broader changes in social structure and environment on individuals' experience of loneliness. Drawing on data from the 2018 Urbanization and Quality of Life Survey (N = 3,229) conducted in 40 localities undergoing rural-urban transition in China, this study investigates how measures of urbanisation (including population density, duration of urban status, neighbourhood transition and housing type) are associated with residents' loneliness. We revised measures of the six-item De Jong Gierveld Loneliness Scale, differentiated between emotional and social loneliness, estimated multi-level mixed-effects regressions and controlled for a number of individual-level covariates. The results show that emotional loneliness and social loneliness have different patterns of association with multi-level covariates: urbanisation at county, township and neighbourhood levels is significantly associated with emotional loneliness, whereas residence in temporary housing is a clear risk factor for social loneliness. The analyses further demonstrate that the revised measures of loneliness address concerns about the original scale, offer a clearer sense of the degrees of loneliness and are strongly associated with multi-level covariates and psychological distress. In addition to showing how urbanisation leads to greater individual loneliness, our research also illustrates how to model locational parameters in analyses of individual well-being.