H. Clark's hypotheses regarding the acquisition of spatial terms were tested using a task designed to elicit subjects' descriptions of different object arrangements. Children from 3 to 10 years and a group of adults were tested. There was some support for Clark's prediction that the perceptual difficulty of each dimension will affect the order of acquisition of the associated linguistic terms, in that the terms followed the predicted order (1) vertical, (2) horizontal-frontal, (3) horizontal-lateral, in the early years. However, it was concluded that since the horizontal-frontal terms lagged behind the others in later childhood and adulthood, perceptual difficulty alone was not an adequate explanation. In addition, no support was found for Clark's notion that linguistic terms associated with the ‘positive’ end of each spatial dimension will be acquired earlier than those associated with the ‘negative’ end.