Saxena et al (Reference Saxena, Levav and Maulik2003) have shown the under-representation of low and middle-income countries on the editorial boards of ten leading psychiatric journals, based on a World Health Organization report. Horton (Reference Horton2003), Editor of The Lancet, has presented some evidence of publication bias against diseases of poverty studied in developing countries. Wilkinson (Reference Wilkinson2003), formerly Editor of the British Journal of Psychiatry, has suggested that the absence of representation on the Editorial Board does not necessarily bias an editor's decision-making. However, Catapano & Castle (Reference Catapano and Castle2003) have shown that research papers from developing countries represent a very small proportion of the publications (<2%) in important psychiatric journals, which has remained the same for 10 years. We argue that Brazil, a middle-income country, is progressively improving its scientific production and reaching the standards of high-income countries.

We have assessed the mental health scientific production of Brazilian postgraduate programmes between 1998 and 2002 using a Brazilian Ministry of Education database. The eight doctoral programmes in psychiatry and psychobiology, all in state institutions, have awarded 183 PhDs and this has resulted in publication of 1664 scientific articles in journals; 605 of these in journals indexed by the Institute of Scientific Information (ISI). The production of ISI-indexed papers doubled in this 5-year-period. The mean impact factor of the ISI-indexed journals where articles were published was 1.82 (range 0.01–29.51); 64% were published in journals with an impact factor41. The number of Brazilian articles in psychiatry and psychology (442) published between 1998 and 2003 corresponds to 10% of France's (4129) production, but the impact factors are very similar: 4.48 and 4.83, respectively (data from ISI, reported on http://in-cities.com/countries).

Although health problems in developing countries account for over 90% of the world's potential life-years lost, only 5% of global health research funds are devoted to these problems (Reference Mari, Lozano and DuleyMari et al, 1997). The investment channelled to postgraduate and human resource educational programmes in Brazil has assured the country a modest but continuous contribution to the worldwide production of knowledge in health. It is expected that the quality of the scientific production of countries such as Brazil will influence editors’ decision-making and overcome eventual ‘institutional racism’ (Reference HortonHorton, 2003).