Abstract This paper analyses monthly hours worked in the US over the sample period 1939m1–2011m10 using a cyclical long memory model. This model, which is based on Gegenbauer processes, is characterised by autocorrelations decaying to zero cyclically and hyperbolically, with a spectral density that is unbounded at a non-zero frequency. One reason for choosing this specification is that the periodogram of the hours worked series has a peak at a frequency away from zero. The empirical results confirm that this model works extremely well for hours worked, and it is then employed to analyse their relationship with productivity. It is found that hours worked increase on impact in response to a technology shock (though the effect dies away rapidly), consistently with Real Business Cycle (RBC) models.