Andersen and Hinthorne reported in situ Pb isotope ages for lunar minerals in the early 1970s. 'Peak stripping' was used to subtract isobaric interferences but it worked only for old, U-rich targets. Jim Long entered the race for in situ mineral dating using high resolution to separate isobars. Despite low sensitivity, he achieved Pb-207/Pb-206 dating of zircons in the late 1970s. In 1980 we finished the large instrument SHRIMP I based on the magnetic analyser design of Matsuda that delivered both high resolution and high sensitivity. A year later we found zircon grains 4200 million years old, far older than any previous terrestrial age. Certain in situ isotope studies remained inaccessible, such as Sr isotopes in Ca-rich minerals. All need 20,000 R with high sensitivity to accumulate enough counts for worthwhile precision. In 1990, Matsuda published new reverse-geometry designs that have still higher resolution, one of which we completed in 1997. We achieved 20,000 R but so far only at reduced sensitivity. Its full potential will be realised after further testing including computer simulations. (c) 2006 Published by Elsevier B.V.