Objective To investigate the correlation of hypomagnesemia and 90-day poor prognosis in the elderly patients with acute heart failure (AHF). Methods A single center prospective trial was conducted on 150 elderly AHF patients admitted in the emergency department of Beijing Hospital between January 2013 and December 2016. According to the occurrence of endpoint events (all-cause mortality and re-hospitalization due to AHF) within 90 d after admission, they were divided into poor prognosis group (n=56) and non-poor prognosis group (n=94). SPSS statistics 19.0 was used to perform the statistical analysis. Student′s t test, Mann-Whitney U test or Chi-square test was employed for comparison between groups. Univariate and multivariate logistic regression analyses were adopted to screen the risk factors influencing the prognosis. Results The incidence rate of 90-day endpoint events were 37.3%(37/150), including 47 cases (31.3%) of re-hospitalization due to AHF and 9 cases (6.0%) of death. The serum magnesium level was significantly lower in the poor prognosis group than the non-poor prognosis group [(0.70±0.05) vs (0.80±0.09) μmol/L, P=0.000]. Multivariate logistic regression analysis indicated that lower magnesium level (OR=30.631,5%CI 5.943-157.881; P=0.000), rapider respiration rate (OR=1.354,5%CI 1.112-1.648; P=0.003), severer grade in bedside classification (OR=3.316, 95%CI 1.169-9.403; P=0.024) and elevated N-terminal pro-brain natriuretic peptide (NT-proBNP) level (OR=3.960, 95%CI 1.318-11.892; P=0.014) were independent risk factors influencing 90-day prognosis of the AHF patients. Conclusion Hypomagnesemia is an independent risk factor for 90-day poor prognosis of elderly AFH patients, and clinicians should pay more attention to its value.