In 14,744 autopsy cases from an 18-year period 92 cases (of which 7 were ruled out because of decomposition were observed in which death was supposed to be due to direct acute alcoholic intoxication. In the police reports 81 persons were designated as chronic alcoholics or abusers of spirits. The blood alcohol level ranged between 2.04 and 4.92 o/oo. The cases studied were divided into two groups, one with low and the other with high lethal alcohol level. Fatty liver and cirrhosis were found with identical frequency in the two groups, whereas cardiac hypertrophy of obscure origin occurred markedly more often in the group with low lethal blood alcohol level. On the basis the possible mechanism of death in the cases with cardiac hypertrophy is discussed. Finally, the relation between the blood and urine alcohol concentrations observed in 72 cases is discussed. On the assumption that the water phase of the blood was 75 per cent of the total blood, death occurred in the persons without cardiac hypertrophy with fairly identical frequency either in the phase of absorption or the phase of elimination, whereas in the persons with cardiac hypertrophy death most often occurred in the phase of absorption. These statements should, however, be taken with some reservation, partly because the water phase of the blood may vary considerably post mortem (60-90 per cent) and partly because the urine alcohol concentration depends on serval variable factors.