Prior research has suggested that among illicit drug users there is an increased risk of drug abuse or dependence problems associated with earlier onset of illicit drug taking. This study examined whether the observed association might be understood best as the result of a process by which earlier onset users accumulate more time during which they can develop a drug problem (that is, more time since first illicit drug taking). Survival data analysis was employed to investigate this issue, with self-report data from 1525 illicit drug using individuals, all of whom were 18-24-year-old participants in the multi-site collaborative Epidemiologic Catchment Area Program. In the first year after initial illicit drug use, there was no appreciable variation in the estimated risk of developing a drug problem. However, with passage of some 5-7 years after initial use, the estimated risk of developing drug problems became clearly greater for youths whose drug taking had started during the pre-adolescent and very early adolescent years. The previously reported relationship between early-onset drug use and subsequent drug problems was not, therefore, an artifact of time. It is possible that duration of drug use should be considered as one 'mechanism' by which early-onset drug users experience greater risk of drug problems, but there also are other interpretations to be investigated in future research.