Data from the Current Population Surveys of October 1984, October 1989, and January 1991 were used to examine the role of computer and mathematical skills in the U.S. labor market from 1984-91. Particular attention was given to their actual and potential effect on the economic status of women. Data confirmed the overall increase in the education level of workers in the U.S. economy. Findings were as follows: college-educated women and men were equally likely to use computer skills on the job; at any other level of education, women were much more likely to use computer skills at the workplace; and the gap between men and women in the rate of computer use widened between 1984-91. The higher rate of computer use among women was a reflection of the continuing segregation of men and women across occupations and industries. Use of computers was highly correlated with use of math on the job. Regression analysis showed that the economy-wide returns to computer use ranged from 13.1 to 18.8 percent in 1991 or about 2 percentage points higher than in 1984 and that, at the level of individual workers, computer skills rewarded men and women equally well. Because women used computers at a higher rate, the gap in average earnings between men and women narrowed slightly with the spread of computers at the workplace. (Appendixes include the following: 24 references, alternative classifications of occupations, job training requirements and skill upgrades, definition of regression variables, and standard errors of parameter estimates.) (YLB)