We have previously reported successful use of a state-of-art home RIP monitor (NIMS, Inc., Miami, FL) as part of the Collaborative Home Infant Monitoring Evaluation (CHIME) study. The current study was performed to develop a model to predict home monitor use in 772 infants enrolled from 5/1/94 to 4/30/97 in 3 groups: siblings of SIDS (Sibs) (n=156), idiopathic apparent life threatening event (n=143), and preterm infants ≤ 34 wks(n=473). Median (25th, 75th%ile) monitor use/wk was 57 (25,94) hrs in wk 1, 53(6,91) in wk 2, 46 (0,80) in wk 3, 36 (0,75) in wk 4 and 32 (0,72) in wk 5. Total use over wks 2-5 was 176 (15,304) hrs. Multiple linear regression analyses were performed on the initial 2 year cohort (n=486) and then prospectively validated on the year 3 cohort (n =286) for prediction of total hours of monitor use in wks 2-5. Variables included: maternal (race, age, marital status, education, parity, smoking/alcohol use during pregnancy), family (no. adults and children in household, self report of family getting along before infant's birth) and infant (sex, birth weight, gestational age, age monitoring began, factors related to preterms: home methylxanthines, ventilator/oxygen days, apnea), and hrs of monitor use in wk 1. When only variables available prior to monitor initiation were used, the model explained only 7% of the variance in monitor use. However, when hrs of monitor use in wk 1 were added, the model explained 62% of the variance. Other significant variables in the final model were race (Black vs. White, reg. coef. -53, p=0.0001), group (Sibs vs. preterms, reg coef. +35, p=0.01), and yrs of maternal education (reg. coef. +4, p=0.03). When this model was prospectively assessed on the year 3 cohort, prediction was comparable to the previous cohort (63% of variance explained). Although characteristics known prior to monitor initiation are not strong predictors of monitor use, knowledge of use during the first week may provide an opportunity to tailor intervention and support of the family.