Until now, compliance with home apnea monitoring (HAM) has been assessed by parental recall. The availability of monitors with electronic memory allows a more objective evaluation of HAM. Fourteen premature infants discharged home from the hospital and receiving theophylline and 15 infants not receiving medication were studied. Medical and birth records, parent questionnaires, and 1 month of electronically recorded data were analyzed. Birth weights of the infants ranged from 850 to 2400 gm and gestational age from 26 to 34 weeks. Of 26 parents who recalled never leaving their infants unmonitored, 70% missed at least one night, 30% three or more. During the first month, treated infants were left unmonitored 21% of the days and 9% of the nights, and untreated babies were unmonitored 32% of the days and 15% of the nights. Average daily HAM use was 18 hours for treated and 16 hours for untreated (p = 0.19). Two weeks after discharge, the decline in use for treated babies was 5% (day and night), whereas for untreated infants the decrease was 20% during the day and 15% at night. Lead-related alarms occurred with similar frequency (one to two times per day) in both groups. Non-lead-related alarms were somewhat more frequent for treated (four times per day) than for untreated infants (twice daily). Parents' perception of the likelihood of severe apnea at the time of discharge influenced HAM use. Little risk prompted 13 hr/day use and moderate risk 16 hr/day, whereas great perceived risk resulted in 18 hr/day of monitoring.(ABSTRACT TRUNCATED AT 250 WORDS)