Objective: To examine the quality of HIV sero-surveillance systems in countries by 2002, as well as trends between 1995 and 2002. Methods: The quality of countries’ surveillance systems was scored for five years: 1995, 1997, 1999, 2001, and 2002. Sero-surveillance data were compiled from the US Census Bureau’s HIV/AIDS Surveillance Database, the EuroHIV database, and from countries’ national HIV surveillance reports that were available to WHO/UNAIDS. The quality of systems was scored according to the level of the countries’ epidemic. Results: There has been a large variation in the quality of HIV surveillance systems across the 132 countries by type of the epidemic and over time from 1995 to 2002. Over the 1995–2002 period the number of countries with a fully implemented system decreased from 57 (43%) in 1995 to 48 (36%) in 2002. The proportion of countries with a fully implemented system was 58%, 34%, and 10% in countries with a generalised, concentrated, and low level epidemic, respectively. In the 53 countries with generalised epidemics the number of countries with a fully implemented system increased from 24 (45%) in 2001 to 31 (58%) in 2002. Conclusion: Many countries still have poor functioning HIV surveillance systems and require urgent strengthening. Countries should monitor and evaluate their own HIV surveillance systems and examine whether the systems are appropriate and adequate.