OBJECTIVES: The transfers of nursing home (NH) residents to the emergency departments (ED) is costly for the healthcare systems. In this regard, the cost-analysis can be a useful tool to provide evidence-based information to guide resource allocation. In this study, we analyze the cost at different stages of the transitions for Edmonton, Alberta, in Canada and evaluate the extent to which nursing home attributes, patients’ characteristics, and the transfers’ outcomes are significantly associated with the transfers cost. METHODS: We used a representative data collected in 2011 from 398 residents and 28 nursing homes in Edmonton, Alberta from multiple levels and sources (facility, residents, healthcare providers, health records, administrative databases, etc.). We conducted a preliminary study to describe the costs at different stages of the transitions. Statistical modeling was used to identify any significant predictors of the transfers cost and estimate the magnitude of their impact. We investigated the ordinary least squares (OLS) models, the GLM, and the mixed effects regression models for clustered data. Models are assessed based on standard goodness-of-fit tests. RESULTS: We find that hospitalization represents 48% of the transfers total cost, whilst the emergency-department nursing cost and the nursing-home cost account for 26% and 10% respectively. The emergency-Medical-Services account for 10%, and the diagnostic cost represents 3%. The physicians plus consultations and the treatment costs represent 2% each. The estimation results show that sociodemographic factors, health characteristics, communication of information, healthcare providers’ involvement in the decision, the nature of the transfer coordinator, and providing medical care in the nursing home are significant predictors of the transfers cost. CONCLUSIONS: The results of this analysis may fit into a subsequent cost-effectiveness analysis for the transfers cost.