The debate about the impact of technology on employment has always had a central role in economic theory. At the same time, the nexus of technological progress and employment might depend on macroeconomic regimes. In this work we investigate the interrelations among technology, output and employment in the U.S. economy in growth recessions vs. growth expansions. More precisely, using U.S. data we estimate different threshold vector autoregressions (TVARs) with TFP, hours, and GDP, employing the latter as threshold variable, and assess the generalized impulse responses of GDP and hours as to TFP shocks. For our entire period of observation, 1957Q1–2011Q4, positive technology shocks, while spurring GDP growth, by and large, display a negative effect on hours worked in growth recessions, but they are not significantly different from zero in good times. Yet, since the mid eighties (1984Q1–2011Q4) productivity shocks increase hours worked in low growth periods. The results are mainly driven by the response of labor along the extensive margin (number of employees), and remain persistent so in the face of a battery of robustness checks.