Many studies of Quaternary climate make use of terrestrial stable isotope recordswhich are interpreted based on seasonal patterns of stable isotopes in modern precipitation. Multi-decade records of isotopes in rainfall allow testing of the assumed behavior of isotope signals used for this interpretation on multi-year to decadal scales. A 32-year record of stable O and H isotopes in precipitation in Tucson, Arizona permits a detailed examination of stable isotope amount effects, at time scales ranging from individual events to decades, in a locationwith summer monsoonal and winter frontal rainy seasons. Amount effects are weak to non-existent in Tucson at seasonal and longer timescales, and are not useful for discriminating either wetter or drier rainy seasons orwetter or drier decades. Amount effects are also weak to non-existent in published data for annual and multi-year amountweighted averages for monsoonal precipitation in New Delhi and Hong Kong, but an annual amount effect appears to be present on Guam (U.S. Territory). In addition, site-specific amount effects do not correlate with measures of regional monsoon intensity. This data analysis challenges the general validity of paleoclimate reconstructions based on short-term (sub-annual) relationships observed in precipitation isotope datawhen applied to long-term records such as speleothem studies.