This study tested the effects of optimized neonatal mother-infant transactions on joint attention performance at 12 months. Surviving infants <2000g from a geographically defined area were randomly assigned to a preterm intervention (n=71) or preterm control group (n=69). Comparisons were made between preterm groups, secondary with a term group (n=75). Joint attention was measured using the Early Social Communication Scales. Preterm intervention infants scored significantly higher than preterm control infants on elements Initiating Joint Attention (p<0.05), Initiating Object Requests (p<0.05), and Responding to Social Interaction (p<0.0005). Intervention was not associated with performance on elements Responding to Joint Attention, or on Responding to Requests. ELBW infants scored significantly lower than VLBW and LBW infants on imperative functions. Girls outperformed boys on all communication elements. An intervention implemented during the neonatal period can be of advantage for certain aspects of joint attention performance in preterm infants.