The effects of microworld microcomputer training on sorting behaviors of 124 twoand three-year-old male and female day care children were studied. Subjects were divided into two age groups (mean ages were 2.5 and 3.0 years) and then into three treatment groups (microworld, real world, and combination) and a control group (no intervention). All treatment groups received one and one-half training hours on an inside/outside a house sorting task using ten familiar, age-appropriate objects. All subjects were pretested and posttested, To assess learning transfer, the posttest included objects on which the children were both trained and untrained. Findings from a 2 x 4 ANCOVA showed a significant age group difference on posttest objects for which children were not trained (p = .0317) and a near significant trend on objects for which the children were trained (p =.0654). Three-year-olds learned better than two-year-olds (p = .0001), with learning increasing over time. One-third of the three-year-olds manipulated the oomputer and task independently. The abstract microcomputer task was shown to be no more difficult for young children than was the concrete doll house task.A number of researchers have called for a period of intensive testing to determine the proper role for the microcomputer in early childhood education [1][2][3]. Data are currently available to show that young children can manipulate the standard microcomputer configuration [4-10], work successfully alone or in small groups [5,8,11,12], manage meaningful microcomputer tasks as young as two to three years of age [8,[12][13][14][15][16], operate and learn from computer