To the Editor.— Arno et al 1 estimate a cost of approximately $5 billion (1989 dollars) for a prospective program of early intervention involving antiretroviral treatment and Pneumocystis carinii pneumonia prophylaxis for human immunodeficiency virus (HIV) pre— acquired immunodeficiency syndrome (AIDS) seropositive patients in the United States. We would like to raise some questions about the validity of two aspects of the authors' calculations, namely, the decline in caseloads for 1989 through 1991 and the adoption of a 50% access-to-treatment level. The General Accounting Office, in its June 1989 report entitled Forecasting, analyzed 13 national forecasts of HIV/AIDS and asserted that a common deficiency of all forecasts was underestimation of the numbers of AIDS cases being projected; a conclusion reached in that report was that most forecasts needed adjustment upward by 50% to compensate for various forms of systematic undercounting. The growth in cumulative AIDS cases is the most