Andreas Stang critically evaluates the Newcastle-Ottawa Scale (NOS) for assessing the quality of non-randomized studies in meta-analyses. The NOS is a tool used to assess the quality of observational studies, consisting of eight items grouped into three dimensions: selection, comparability, and outcome (for cohort studies) or exposure (for case-control studies). It uses a star system to assign a semi-quantitative score, with a maximum of nine stars. However, Stang argues that the NOS has several issues, including unclear validity and potential bias in its scoring system. In case-control studies, the NOS gives higher scores to studies with independent validation of case status, but this may not always reflect true validation. It also assigns higher scores to studies with hospital-based controls, despite the lack of a general preference for community-based controls. Additionally, the NOS gives higher scores to studies with comparable non-response rates between cases and controls, but this may not account for selection bias. Stang points out that identical response rates between cases and controls do not necessarily eliminate selection bias. In cohort studies, the NOS gives higher scores to studies with community-representative exposed cohorts, but this may not always lead to better generalizability. It also assigns the same score to studies with independent or blinded outcome assessments or record linkage, which may not reflect the true quality of the study. Stang concludes that the NOS may not be a reliable tool for assessing the quality of non-randomized studies in meta-analyses, as it may include items with uncertain validity and may lead to arbitrary results. The use of the NOS in evidence-based reviews and meta-analyses may therefore be problematic.Andreas Stang critically evaluates the Newcastle-Ottawa Scale (NOS) for assessing the quality of non-randomized studies in meta-analyses. The NOS is a tool used to assess the quality of observational studies, consisting of eight items grouped into three dimensions: selection, comparability, and outcome (for cohort studies) or exposure (for case-control studies). It uses a star system to assign a semi-quantitative score, with a maximum of nine stars. However, Stang argues that the NOS has several issues, including unclear validity and potential bias in its scoring system. In case-control studies, the NOS gives higher scores to studies with independent validation of case status, but this may not always reflect true validation. It also assigns higher scores to studies with hospital-based controls, despite the lack of a general preference for community-based controls. Additionally, the NOS gives higher scores to studies with comparable non-response rates between cases and controls, but this may not account for selection bias. Stang points out that identical response rates between cases and controls do not necessarily eliminate selection bias. In cohort studies, the NOS gives higher scores to studies with community-representative exposed cohorts, but this may not always lead to better generalizability. It also assigns the same score to studies with independent or blinded outcome assessments or record linkage, which may not reflect the true quality of the study. Stang concludes that the NOS may not be a reliable tool for assessing the quality of non-randomized studies in meta-analyses, as it may include items with uncertain validity and may lead to arbitrary results. The use of the NOS in evidence-based reviews and meta-analyses may therefore be problematic.