Background: Atraumatic splenic rupture (ASR) is an ill defined clinicopathological entity. Methods: The aim was to characterize aetiological and risk factors for ASR‐related mortality in order to aid disease classification and treatment. A systematic literature review (1980–2008) was undertaken and logistic regression analysis employed. Results: Some 632 publications reporting 845 patients were identified. The spleen was normal in 7·0 per cent (atraumatic–idiopathic rupture). One, two or three aetiological factors were found in 84·1, 8·2 and 0·7 per cent respectively (atraumatic–pathological rupture). Six major aetiological groups were defined: neoplastic (30·3 per cent), infectious (27·3 per cent), inflammatory, non‐infectious (20·0 per cent), drug‐ and treatment‐related (9·2 per cent) and mechanical (6·8 per cent) disorders, and normal spleen (6·4 per cent). Treatment comprised total splenectomy (84·1 per cent), organ‐preserving surgery (1·2 per cent) or conservative measures (14·7 per cent). The ASR‐related mortality rate was 12·2 per cent. Splenomegaly (P = 0·040), age above 40 years (P = 0·007) and neoplastic disorders (P = 0·008) were associated with increased ASR‐related mortality on multivariable analysis. Conclusion: The condition can be classified simply into atraumatic–idiopathic (7·0 per cent) and atraumatic‐pathological (93·0 per cent) splenic rupture. Splenomegaly, advanced age and neoplastic disorders are associated with increased ASR‐related mortality. Copyright © 2009 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd. Usually associated with underlying pathology