17561 Background: The science and practice of cancer medicine in Brazil currently display features of both developing and developed countries. Cancer research is increasing in Brazil, and several recent large, international, industry-sponsored trials have included a substantial number of Brazilian patients. However, it is unclear to what extent there is proportional increase in research originated in Brazil. We approached this question using a bibliometric analysis of Brazilian abstracts from recent ASCO Annual Meetings. Methods: We conducted a manual search of the 16,925 abstracts published in the Program Proceedings and late-breaking abstracts booklets for 2001 through 2005. We defined as Brazilian an abstract for which at least two-thirds of institutions were from Brazil. We tabulated study characteristics and searched PubMed and Lilacs databases for their full publications, using first and last authors’ names. Results: We located 154 Brazilian studies (0.90% of total; 95% CI 0.77% - 1.07%), of which 76.0% were categorized as clinical investigation, 14.9% as basic research, and 9.1% as epidemiologic research. In clinical investigation, the most frequent study types were retrospective (N=43) and phase II trials (N=37). The most frequent topic categories were tumor biology (N=27), patient care (N=22), breast (N=19), lung and gastrointestinal cancer (N=13 each). There was a significant increase in the proportion of Brazilian abstracts from 2001 to 2005 (P=0.011 for trend). Only 31.8% of the studies were presented at the meeting, usually as posters, with the other 68.2% appearing as “publication only”. There was no significant trend in the proportion of “publication only” abstracts (P=0.407). After a median follow up of 35 months, only 16.9% of studies were published (median time to publication, 13.5 months). Univariate Kaplan-Meier analyses showed no significant predictors of publication, although “publication only” abstracts were less likely to be published in full (P=0.059). Conclusions: Our study provides empiric evidence for an increase in research that is originated in Brazil. However, it also suggests that the quality of such research needs to improved, and that efforts are also needed in order to increase the rate of full publication. No significant financial relationships to disclose.