BACKGROUND: The availability of medicines on the illicit drug market is currently high on the international policy agenda, linked to adverse health consequences including addiction, drug related overdoses and injection related problems. Continuous surveillance of illicit use of medicines allows for earlier identification and reporting of emerging trends and increased possibilities for earlier intervention to prevent spread of use and drug related harm. This paper aims to identify data sources capable of monitoring the illicit use of medicines; present trend findings for Rohypnol and Subutex using a multi-indicator monitoring approach; and consider the relevance of such models for policy makers. METHODS: Data collection and analysis were undertaken in Bergen, Norway, using the Bergen Earlier Warning System (BEWS), a multi-indicator drug monitoring system. Data were gathered at six monthly intervals from April 2002 to September 2006. Drug indicator data from seizures, treatment, pharmacy sales, helplines, key informants and media monitoring were triangulated and an aggregated differential was used to plot trends. RESULTS: Results for the 4-year period showed a decline in the illicit use of Rohypnol and an increase in the illicit use of Subutex. CONCLUSION: Multi-indicator surveillance models can play a strategic role in the earlier identification and reporting of emerging trends in illicit use of medicines.