This paper explores the interplay of maternal sensitivity (or, more generally, the quality of the caregiving social environment) and infant individual dispositions in predicting infant-mother attachment. After a brief theoretical introduction, the focus turns to studies conducted during the 1980s that predicted attachment security vs. insecurity at 12 months from newborns' ability to regulate orientation and arousal. A re-analysis of two longitudinal studies, formerly coded only with the ABC system, subsequently revealed that disorganized (vs. organized) attachment was predictable from newborns' regulatory abilities, whereas secure (vs. insecure) attachment was predictable from the quality of maternal care. This suggested that the two dichotomies represented distinct dimensions and that - in low risk samples - disorganized SSP behavior may be associated with infant behavioral dispositions. More recent attempts to predict disorganized attachment from infant genetic (as opposed to newborn behavioral) dispositions yielded inconsistent results when only main effects models were examined. A set of subsequent studies examining the interplay of genetic and caregiving influences in the prediction of disorganized attachment suggested that maternal sensitivity is of particular importance in cases of genetic risk.