Using data from the first six waves of the British Household Panel Survey, we estimate the impact of working longer hours over 1991 to 1995 on 1996 wages. We find that there are positive but diminishing long-term returns, with the returns becoming negative beyond 47 hours for women and 59 hours for men. The returns are greater at the margin for unpaid hours than for paid hours. Evaluated at the mean, an extra unpaid hour over 1991 to 1995 raised 1996 pay by 4 percent, an extra paid hour by 1 percent. It also pays off to work longer hours than norm for the industry. While there are no significant differences between the marginal effects for men and women, conditional on hours worked the incentives are greater for women than for men. These findings are consistent with the possibility that increasing UK wage inequality is associated with an upward impact on work hours.