The literature on endogenous growth cycles predicts countercyclical R&D expenditure. Aggregate R&D expenditure in G7 countries from 1973 to 1997 seems to be procyclical. Implications for future theoretical research are discussed.