Several types of job-analysis inventories or instruments are examined in terms of their utility for skill forecasting. It is argued that task and skill-ability job-analytic information may be useful when task information is used to make inferences concerning which skills-abilities will be necessary in future jobs. An illustration is presented in which a task and ability oriented job-analysis inventory was completed by 371 incumbents in nine skilled-trades occupations to form a matrix of intercorrelations between specific task and skill-ability characteristics. Multiple regression methods were subsequently applied given assumptions of different numbers of tasks being hypothetically available to decision makers to predict which skills-abilities would be necessary in some future job. The decision outcomes formed from using these regression equations were cross validated with an additional 248 incumbents. Results indicated that such forecasting predictions can represent useful information to decision makers. Discussion centers on the implications and limitations of this forecasting procedure.