Background The acceptance of a paper in a top-ranked journal depends on the importance of the study, and should not depend on its country of origin. If the papers' citation rate is a proxy for their importance, and the threshold for acceptance is unrelated to the country of origin, papers from different countries published in the same journal should have a similar number of citations. Conversely, if the threshold is lowered for some countries, their papers will have a lower mean citation rate.

Methods The number of citations and the corresponding author's country were obtained for 4724 papers published between 1998 and 2002 in the British Medical Journal, the Lancet, Journal of the American Medical Association and New England Journal of Medicine. Countries were grouped according to the World Bank classification and geographical location: low–middle income countries; high-income European countries; high-income non-European countries; UK and USA. The probability of papers being poorly cited by country of origin was estimated, using domestic papers (British papers published in British journals and US papers published in US journals) as the reference.

Results Compared with domestic papers, the OR of being poorly cited was 0.67 (95% CI 0.55 to 0.81) for papers from high-income European countries, 0.97 (0.76 to 1.24) for papers from high-income non-European countries and 1.93 (1.28 to 2.89) for papers from low–middle income countries.

Conclusions Papers from different countries published in the same journal have different citation rates. This may reflect difficulties for researchers from some countries to publish their research in leading medical journals.