This paper reports on the development and validation of an instrument for recording haptic exploratory behavior through detailed examination of exploratory search strategies of young children. Individual differences within age level were investigated, as well as the extent to which information processing capabilities for haptic perception are dependent upon developmental level, the stimulus properties of particular shapes, and hemispheric specialization. Subjects were 76 school children aged 53 to 89 months who were tested individually on identification of 3-dimensional tactual stimuli in the absence of any visual cues. A haptic perception box, with holes for insertion of hands, allowed the child to explore pairs of shapes (identical or different) in a series of tests involving five types of shapes: topological (familiar or unfamiliar) or Euclidean (curvilinear or rectilinear), or a combination of both. Haptic exploration and accuracy, highly correlated with traditional tests of spatial ability, were found to be clearly dependent upon the child's developmental level. Significant intra-individual and inter-individual differences were noted. Success on a task was found to be a complex function dependent upon developmental level, task requirements, procedural variations, stimulus properties and individual abilities, as well as on the concomitant interactions. Evidence in support of Gibson's theory of distinctive features is provided. Evidence on cerebral dominance was inconclusive. (BF)