This study evaluates children's performance on selected spatial prepositions and determines the age levels these prepositions are acquired in both receptive and expresive language, as revealed in tasks involving both two- and three-dimensional objects. Subjects were 80 children (40 males and 40 females), ranging in age from three years to four years and eleven months. All were native English speakers with no speech, hearing, or neurological disorders, and with normal intelligence. Results indicated a significant difference in test scores according to age (older children perform better than younger), task (comprehension scores higher than production scores), referent (three-dimensional tasks showing higher scores than two-dimensional tasks), and preposition. Children's use of selected spatial prepositions is dependent on the semantic complexity of the preposition. Prepositions whose meanings can be described in terms of simple topological notions are understood and used with greater facility than those involving dimensional or Euclidean spatial notions. When the prepositional variable interacts with age, dimension, task, age + dimension, age + task, dimension + task, and age + dimension + task, overall differential response are likely to occur.