Background: Previous studies have shown that substance use among homeless people is a prevalent problem that is associated with longer durations of homelessness. Most studies of substance use among the homeless were carried out outside Europe and have limited generalizability to European countries. This study therefore aimed to address the prevalence of substance use among homeless people in the Netherlands, the pattern of their use and the relationship with housing status at follow-up. Methods: This study included 344 participants (67.1% of the initial cohort) who were followed from baseline to 18 months after the baseline interview. Multinomial logistic regression analyses examined the relationship between substance use and housing status. Results: The most reported substances which were used among these homeless people were cannabis (43.9%) and alcohol (>= 5 units on one occasion) (30.7%). Other substances were used by around 5% or less of the participants. Twenty-seven percent were classified as substance misuser and 20.9% as substance dependent. The odds to be marginally housed (4.14) or institutionalized (2.12) at follow-up compared to being housed of participants who were substance users were significantly higher than those of participants who did not use substances. The odds to be homeless were more than twice as high (2.80) for participants who were substance dependent compared with those who were not. Conclusion: Homeless people who use substances have a more disadvantageous housing situation at follow-up than homeless people who do not use substances. Attention is needed to prevent and reduce long-term homelessness among substance-using homeless people.