This paper adresses the question wether patients have a higher risk of death if they are admitted to a hospital during the night rather than during the day. Since hospital resources and stuffing are reduced at nights, both numerically and with regard to available expertise on site, there is a theoretical channel, explaining why night admissions might be more dangerous. However, since patients admitted during the night are special in terms of urgency and admission time, simple comparisons of the groups may be misleading. To facilitate comparison, the empirical analysis compares patients who suffer from specific deseases considered as emergency conditions, i.e. upper gastrointestinal bleeding, or acute myocardial infarction and is based on regression adjusted propensity score matching as identification strategy. A full sample of all inpatients in Germany for the years 2005 to 2007 is used. Additionally, I execute two sensitivity analyses to determine how unobserved heterogeneity influences the results. Although the results show that patients admitted during the night have a higher risk of death, sensitivity analyses suggest, that this result is likely to be driven by unobserved patient heterogeneity.