INTRODUCTION. To improve the outcomes of dyspnoeic patients, it is potentially important to determine the influence of uncertain diagnostic aetiology and to characterise this patient group. A change in diagnosis from emergency department (ED) contact to hospital discharge (i.e., a discordant diagnosis) served as a surrogate measure for an uncertain diagnostic aetiology. This study investigated the association between a change in diagnosis from ED contact to hospital discharge and length of stay (LOS), readmission and mortality in patients whose chief complaint was dyspnoea. The study also characterises the group of patients found to have a discordant diagnosis. METHODS. This cohort study was based on data from all ED contacts at Aarhus University Hospital from 1 July 2016 to 30 June 2017. Patients triaged with dyspnoea and subsequently admitted to an inpatient unit were included. RESULTS. Concordant contacts had an average LOS of 3.63 days, whereas discordant contacts had an average LOS of 4.65 days; the adjusted relative difference was 1.28 (95% CI: 1.10-1.48). Readmission,  whether at seven or 30 days, was not significantly different between the groups. The 30-day mortality was 5% in the concordant and 10% in the discordant group, with an adjusted OR of 2.32 (95% CI: 1.08-4.96). CONCLUSIONS. We found an association between a change in diagnosis and longer LOS, and between a change in diagnosis and 30-day mortality. The effort made to achieve diagnostic certainty in the ED may have an impact lasting throughout the entire hospital stay.