Two experiments investigated preschoolers' acquisition of spatial words in Mandarin Chinese. In one experiment, 5 groups of 10 children at 34, 39, 46, 52, and 57 months were tested for comprehension and production of 14 pairs of Chinese spatial words. In the comprehension test the children were asked to point to pictures corresponding to the words spoken; in the production test they were asked to say the opposite of the word spoken by the experimenter. Results indicated a high degree of consistency in the order of acquisition, which paralleled findings in English acquisition studies and was consistent with theory. However, the theory that unmarked words should be learned before the marked words was not supported. In the second experiment 3 groups of 10 children aged 34, 46, and 57 months were tested for their comprehension of two pairs of spatial terms. An object was placed in relation to another object having one or the other of the following characteristics: (1) inherent top, bottom, front, and back; (2) inherent front/back but not top/bottom; (3) inherent top/bottom but not front/back; and (4) neither top/bottom nor front/back. Results indicated that under all conditions even the youngest children had almost perfect comprehension of top and bottom but imperfect comprehension of front/back for objects without inherent front/back markings. (MSE)