The Swedish National Inpatient Register (IPR), also known as the Hospital Discharge Register, is a primary source of data for numerous research projects. Launched in 1964 and fully operational since 1987, the IPR covers over 99% of all somatic and psychiatric hospital discharges in Sweden. The register is part of the National Patient Register and uses the Swedish International Classification of Diseases (ICD) system for coding diagnoses. A previous validation by the National Board of Health and Welfare showed that 85-95% of diagnoses in the IPR are valid. This paper reviews the history, structure, coverage, and quality of the Swedish IPR. The review involved searching medical databases and contacting experts to identify 132 papers that validated the IPR. The positive predictive value (PPV) of diagnoses in the IPR was found to be 85-95%, with variations depending on the diagnosis. The IPR is particularly suitable for large-scale population-based research due to its long follow-up period, but for certain research areas, other health registers like the Swedish Cancer Register may be more appropriate. The study concludes that while the IPR has high validity for many diagnoses, it may have lower sensitivity for certain conditions, such as hypertension and lipid disorders. The introduction of outpatient care and financial incentives for coding certain diagnoses may have influenced the coding patterns and affected the epidemiological patterns. Despite these limitations, the Swedish IPR remains a valuable resource for research.The Swedish National Inpatient Register (IPR), also known as the Hospital Discharge Register, is a primary source of data for numerous research projects. Launched in 1964 and fully operational since 1987, the IPR covers over 99% of all somatic and psychiatric hospital discharges in Sweden. The register is part of the National Patient Register and uses the Swedish International Classification of Diseases (ICD) system for coding diagnoses. A previous validation by the National Board of Health and Welfare showed that 85-95% of diagnoses in the IPR are valid. This paper reviews the history, structure, coverage, and quality of the Swedish IPR. The review involved searching medical databases and contacting experts to identify 132 papers that validated the IPR. The positive predictive value (PPV) of diagnoses in the IPR was found to be 85-95%, with variations depending on the diagnosis. The IPR is particularly suitable for large-scale population-based research due to its long follow-up period, but for certain research areas, other health registers like the Swedish Cancer Register may be more appropriate. The study concludes that while the IPR has high validity for many diagnoses, it may have lower sensitivity for certain conditions, such as hypertension and lipid disorders. The introduction of outpatient care and financial incentives for coding certain diagnoses may have influenced the coding patterns and affected the epidemiological patterns. Despite these limitations, the Swedish IPR remains a valuable resource for research.