IN A RECENT issue of this Journal, Fortune [1] published a paper purporting to provide a theoretical foundation for the analysis of optimal expenditure on life insurance. The purpose of this comment is to show that Fortune's model has certain formal implications which are so at variance with conventional assumptions about human preferences as to detract from the relevance of his analysis and conclusions. Briefly, Fortune considers a two-period situation in which an individual will die either in period 1 or in period 2. Without life insurance the present value of the individual's estate is W1 if he does not survive period 1 and W2 if he survives to period 2. The individual may, however, enter into a life insurance contract whereby he pays a premium P in order that the present value of his estate will be augmented by an amount INS1 should he die in period 1 or INS2 if he survives period 1 and dies in period 2. Fortune's basic choice postulate is that the individual is an 'expected utility maximizer' so that he will purchase life insurance if and only if doing so results in an increase in expected utility. More formally, Fortune's condition for life insurance to be purchased is that