Theoretical predictions of the impact of TFP growth on unemployment are ambiguous, and depend on the extent to which new technology is embodied in new jobs. We evaluate a model with embodied and disembodied technology, capitalization, and creative destruction effects by estimating the impact of TFP growth on unemployment in a panel of industrial countries. We find a large negative impact which implies that embodied technology and creative destruction play no role in the steady-state dynamics of unemployment. Capitalization effects explain some of the estimated impact but a part remains unexplained.