The history of PET research clearly demonstrates that it is advances in chemistry coupled with a detailed examination of the biochemistry of new radiotracers that have allowed the PET method to be applied to new areas of biology and medicine. Radiotracers whose regional distribution reflects glucose metabolism, neurotransmitter activity and enzyme activity have all required the development of rapid synthetic methods for the radiotracers themselves and the characterization of their biochemical behavior. This article traces some of the advances in the production of labeled precursors and in radiotracer synthesis and evaluation which have shaped the rapidly expanding application of PET to problems in the neurosciences, in cardiology and in oncology.