We examined whether a maximum threshold of time spent in nonmaternal care exists, beyond which infants have an increased risk of forming a disorganized infant-mother attachment. The hours per week infants spent in nonmaternal care at 7-8 months were examined as a continuous measure and as a dichotomous threshold (over 40, 50 and 60 hr/week) to predict infant disorganization at 12-15 months. Two different samples (Austin and NICHD) were used to replicate findings and control for critical covariates: mothers' unresolved status and frightening behavior (assessed in the Austin sample, N=125), quality of nonmaternal caregiving (assessed in the NICHD sample, N=1,135), and family income and infant temperament (assessed in both samples). Only very extensive hours of nonmaternal care (over 60 hr/week) and mothers' frightening behavior independently predicted attachment disorganization. A polynomial logistic regression performed on the larger NICHD sample indicated that the risk of disorganized attachment exponentially increased after exceeding 60 hr/week. In addition, very extensive hours of nonmaternal care only predicted attachment disorganization after age 6 months (not prior). Findings suggest that during a sensitive period of attachment formation, infants who spend more than 60 hr/week in nonmaternal care may be at an increased risk of forming a disorganized attachment. (author abstract)