Spatial relations in American Sign Language (ASL) are often signed from the perspective of the signer and so involve a shift in perspective and mental rotation. This study examined developing knowledge of language used to refer to the spatial relations "front," "behind," "left," "right," "towards," "away," "above," and "below" by children learning ASL and English. Because ASL is a classifier language in which noun referents are placed into groups, each spatial relation also appeared with person, animal, and vehicle classifiers. Twenty-three children and adults who learned ASL before the age of 5 years and 23 native English-speaking adults and children participated. Both language groups participated in a comprehension task in which they chose which of 2 pictures depicted a signed or spoken relation. Results showed that children learning ASL acquired the constructions for spatial relations that typically involve perspective shifts and mental rotation later than constructions that do not involve these abilities and later than English-speaking children. Children learning ASL did not differ from English-speaking children in learning constructions that did not involve these abilities. Results also suggest that users of ASL initially comprehend spatial relations more accurately with person and animal classifiers than with the classifier for vehicles. The results are relevant to understanding the acquisition of spatial relations in ASL.