Infants' cross-modal functioning was investigated in two studies. In Study 1, 11-month-old infants were confronted with five different visual-tactual discrepancies created with a mirror arrangement. The infants' behavioral reactions to the discrepancies were compared with their behavior on matched control trials with a forced-choice judgement procedure. Infants detected discrepancies in which they saw an egg and felt a cube, saw a fur-covered cube and felt an egg, and saw a cross and felt a fur-covered cube. However, they provided no evidence that they detected discrepancies in which they saw a cube and felt a cross or saw a cube and felt a fur-covered cube. In Study 2, infants were confronted with discrepancies that were the converse of those which seemed to go unnoticed in Study 1: They saw either a cross or a fur-covered cube and felt a plain cube. Both of these new discrepancies were detected according to the forced-choice judgment procedure. The results indicate that texture as well as shape can serve as a basis for cross-modal matching for infants. The asymmetries in cross-modal matching that were observed across Studies 1 and 2 are interpreted as evidence that visual information plays a directive, goal-setting role for infants' manual explorations.