To the Editor.Ìý —In the April 1986 issue of the Archives, Melamed¹ and Markham and Diamond² argue the merits of delaying levodopa therapy in parkinsonism. As they so well demonstrated, an absolute conclusion is precluded by the extreme complexity of the disease and the difficulty in developing studies that unequivocally resolve the issue. Nevertheless, there is a factor, not mentioned, that definitively settles the controversy, that of risk aversion.³ All of us are averse to taking risks, and the degree of aversion increases the sooner the risk is to be taken. A corollary is that the value of benefits is enhanced if they can be enjoyed immediately and reduced the longer they are delayed. For the parkinsonian patient, the value of a disability-free year now is likely to be several times the value of a disability-free year five years hence. The exact relative value, r, can be easily