Hospital admission trends in Myasthenia Gravis are largely unknown, so they were here investigated in Finland between 2004 and 2014 using national mandatory registry data. There were 2989 hospital admissions (59.7% for women) for 861 individuals (median 2 admissions/individual) The annual number of admissions (p = .56), the age of admitted patients (p = .24) or length of stay (p = .20) showed no change during the study period. The proportion of infections as the primary diagnosis increased from 4.5% to 10.4% (p = .0056). These admissions lasted longer than admissions with a non-infectious primary diagnosis (median 6 vs. 4 days, p < .0001). In-hospital mortality rate was 1.0%, predicted by age over 65 (HR 8.8; p = .0034) and infection as the primary diagnosis (HR 6.9; p < .0001). Annual frequencies of thymectomies (p = .66) or plasmaphereses (p = .12) remained unchanged. Myasthenia drug reimbursement data suggested increasing MG prevalence during the study period (p < .00001). Considering that the annual hospitalisation frequency remained stable, this would suggest decreased need of hospitalisations per patient. The importance of infections as causes of myasthenia hospitalisations merits further study.