In Health Economics and Outcomes Research most statistical analyses are embedded in the field of epidemiology or econometrics. However, there is still a gap between the two fields in terms of the used key terminology for statistical concepts. Amongst others, this gap is reflected in the different interpretations of the concept of confounding, and related terms such as selection bias. The problem of confounding appears often in biomedical research. Econometricians often refer to this phenomena as an endogeneity problem, omitted variable bias or unobserved heterogeneity, whereas epidemiologists typically use the term confounding by indication. Furthermore, if we compare the use of the term selection bias within the two disciplines, we can see that selection bias in epidemiology is most commonly used for describing the situation when the study population is not comparable to the source population. For example, if randomized controlled trial participants are healthier, the observed sample is not representative of the source population. In econometrics, however, the term selection bias is mainly used when the treatment uptake is non-random due to self-selection of individuals. This may in turn result in self selection bias when there is a comparison between hospitalized and non-hospitalized individuals without taking into account that hospitalized individuals tend to be less healthy. This poster presentation aims to shed light on the fine differences in terminology for confounding, and its associated terms, within and between each field. We provide examples in order to emphasize the minor differences between apparently similar concepts. We hope that by discerning the difference in approaching methodological issues between the disciplines, a better communication and cooperation between them is stimulated.