This paper investigates the impact of computer usage at work and other job features on the changing skills required of workers. It compare skills utilization in Britain at three data points: 1986, 1992 and 1997, using proxies for the level of skills actually used in jobs. This study questions the validity of investigating the facts about, and the sources of, rising skills by using just educational attainment or occupational grouping data. This paper finds that: ¢ Job skills have increased, more quickly for women than for men; these increases are more extensive than those captured by changes in the occupational class structure. ¢ The spread of computer usage is very strongly associated with the process of upskilling, and accounts in part for narrowing of the gender skills gap; expanded use of quality circles is also linked to upskilling. Evidence for any direct role of trade in inducing skills increases is weak. ¢ Using the qualification held or occupation as a skills measure can lead to erroneous conclusions as to the origin of skills changes.