Infant sleep disturbance has been effectively treated with extinction by removal of parental attention but this procedure can produce adverse side-effects and can meet with parental resistance. This study evaluated an alternative procedure using graduated extinction to treat infant sleep disturbance. Graduated extinction required parents to gradually reduce the amount of time they spent attending to their child, eventually withdrawing attention completely. Baseline measures of sleep-related behaviour, including frequency and duration of night-waking, sleep onset latency and bedtime delay, were made by parents for twelve children between 6 and 19 months. Seven children received graduated extinction, and five received direct extinction in a multiple-baseline across subjects design. Social validation measures were taken during baseline and at the end of treatment, and sleep-related behaviours were again assessed for one week at two-months follow-up. Clinically significant improvements in infant sleep behaviour were evident for four out of six subjects exposed to graduated extinction and for three out of four given extinction. Data for two subjects were unavailable for analysis. Improvements were maintained at follow-up. Parents reported high satisfaction with the procedures and improved family lifestyles. The quality of parental compliance with treatment was lower for graduated extinction than for extinction. The conclusions revealed that while graduated extinction was effective in treating infant sleep disturbance, it is no more effective than extinction, which has practical and clinical advantages in almost all cases.