The pr edi cti ve r el ati on betw een neonatal r eacti vi ty and 12-, 18-, and 24-month temper ament w as assessed for 22 appr opr i ate-for-gestati onal -age (A GA )/smal l -for-gestati onal -age (SGA ) tw i n pai r s. Neonatal r eacti vi ty i ncl uded r ati ngs of vi sual and audi tor y or i enti ng r esponses to a bul l 's-eye, r attl e, bel l , voi ce, and face pl us voi ce combi ned, as w el l as al er tness. I nfant temper ament r ati ngs w er e made of emoti onal tone, acti vi ty, attenti veness, and soci al or i entati on to staff dur i ng ageappr opr i ate acti vi ti es i n the l abor ator y. Ther e w er e no gr oup di ffer ences i n the neonatal or i nfant r ati ngs. For A GA tw i ns, a patter n of pr edi cti ve cor r el ati ons w as obtai ned betw een neonatal r eacti vi ty and temper ament at each age. A si mi l ar pr edi cti ve patter n w as not obser ved for the SGA tw i ns. Gr oup di ffer ences i n r ates of matur ati on r el ated to str esses associ ated w i th i ntr auter i ne gr ow th r etar dati on i nfl uenced the si gni fi cance of the neonatal var i abl es for l ater temper ament devel opment. Keyw or ds: tw i ns, si ze-for-gestati onal -age, neonate, i nfant, temperament, devel opment, ori enti ng, i ntrauteri ne grow th retardati on