Stereopsis was assessed in 94 children ages 2–5 yr using random-dot stereograms (RDSs) and operant conditioning. After five traditional screening tests of stereopsis, all children were trained to perform a match-to-sample discrimination task involving RDSs with and without lateral disparity. An added feature of the operant RDS task was the availability of monocular cue-fading for children who failed initial RDS testing. The operant RDS test facilitated stereopsis-based responding more than the traditional clinical tests did. This finding is interpreted as indicating the importance of such factors as attention, motivation, and language ability in the assessment of stereopsis in young children.