Traditional management of juvenile diabetes involves insulin doses high enough to prevent any trace of sugar or acetone in the urine, and dietary carbohydrates are strictly limited lest they show up as urinary sugar. Lowenstein and Preger find this approach not only inadequate but dangerous. They cite recent suggestions that the well-known progressive ""complications"" of diabetes result largely from insulin, that urinary sugar of itself never killed anyone, and that insulin dosage should be held to the absolute minimum necessary to suppress acetone. They think the low-carbohydrate diet (useful in the kindred condition of hypoglycemia) actually increases the insulin requirements of true diabetics. More fundamentally, they note a shift of opinion on what diabetes is. The supposed vascular ""complications"" (deposits of ""metabolic debris"" on the walls of tiny blood vessels, causing ever-worsening circulatory problems of every sort from retinal hemorrhage to impotence to gangrene) now seem to be an intrinsic part of the disease from the preclinical stage--perhaps even the cause of the ""primary"" metabolic derangement. At any rate intravenous insulin aggravates arteriosclerosis--a necessary evil until transplants of insulin-producing cells become feasible. Meanwhile the authors recommend a low-fat, moderately high-carbohydrate diet, conservative insulin management (there are discussions of the various types now in use), and--for those with ""adult"" diabetes--avoidance of all oral insulin substitutes. The treatment is often unfocused; there are annoying repetitions and occasional excursions into fashionable trivia. But the main outlines of the argument are solidly persuasive.