Our objective was to estimate the saving of life by thrombolysis and resuscitation in acute myocardial infarction (AMI) before and after hospital admission. We studied all 1516 patients admitted to a Scottish teaching hospital in 1990 and 1992 who had a final diagnosis of AMI, and 311 patients enrolled in the Grampian region early anistreplase trial (GREAT). Cardiac arrest occurred in 250 1516 (16%) hospital patients. Of these, 77 (31%) were discharged alive — a saving of 51 lives per thousand cases. 797 (53%) patients received thrombolysis, of whom 114 (14%) died. Assuming the same relative reduction in mortality as in the second international study of infarct survival (ISIS-2; 23%), 34 lives were saved by thrombolytic therapy, representing 22 lives per thousand cases. Of 311 patients in GREAT, 15 (5%) had prehospital cardiac arrest, with 7 patients surviving to discharge (48%) — a saving of 23 lives per thousand cases. Those patients given domiciliary thrombolysis had a one month mortality of 6.7% ( 11 163 ) compared with 12.2% ( 18 148 ) for those receiving hospital thrombolysis — a saving of 55 lives per thousand cases for prehospital thrombolysis. This is additional to 28 lives per thousand estimated for thrombolytic therapy in hospital, totalling 83 lives saved per thousand cases of AMI receiving thrombolytic therapy prehospital. In hospital, more lives were saved by resuscitation than by thrombolytic therapy, but this ratio was reversed in the period before hospital admission. These results emphasise the paramount importance of resuscitation in hospital, and the enhanced efficacy of thrombolysis when given at the earliest opportunity.