The role of decision making and keystroking in the psychomotor activity of office typing tasks was assessed for second-semester, fourth-semester, and terminal college students under three different work conditions: (1) pre-arranged, (2) unarranged without erasing errors, and (3) unarranged with erasing errors. All differences for main effects for speed and errors were significant (probability less than .01). When office tasks were done under wholly realistic conditions, planning and decision making were one-half, keystroking was three-eights, and erasing comprised one-eighth of the task. Mean completion time for one letter, one table, and one rough draft for 60 students was: prearranged without erasing, 8.94 minutes; unarranged without erasing, 20.89 minutes; and unarranged with erasing, 23.93 minutes. At low levels of skill, time consumed for keystroking was nearly as salient in contributing toward completion of product as time consumed for decision making. As psychomotor skill increased, the perceptual skill of decision making played an increasingly dominant role. The amount of time needed for decision making took on increasing salience in producing office communications as difficulty of the task and amount of training were increased. (Author)