Background: Japanese biomedical researchers are said to have been using less sophisticated study designs in terms of clinical research, and studies on quantitative estimates based on randomized controlled trials (RCT), cohort studies (CS), case-control studies (CCS) and meta-analyses (MA) are not available in this regard to date. Methods: We searched PubMed database to estimate the number of RCT, CS, CCS and MA published by Japanese researchers during the period 1994-2003, and compared it with that of rest of the countries as a whole. The search criteria we employed included publication year (1994-2003), language (English), tag (human), publication type (MA, RCT), and medical subjects headings (cohort studies, case-control studies, odds ratio). Nonparametric tests for trends were performed to determine any significant changes in the number and proportion of publication types over the period in question. Results: During the period 1994-2003, 98,774 RCT, 174,898 CS, 5,489 CCS, and 6,993 MA were published as a whole while Japanese researchers contributed 3,148 (3.19%), 8,985 (5.36%), 268 (4.03%), and 62 (0.89%) papers respectively. For the year 1994, the respective percentages were 2.58%, 4.02%, 2.25%, and 0%, and for the year 2003 they were 3.39%, 5.36%, 5.49%, and 1.15%. Japan's contribution has increased significantly over the period of time, in all categories in absolute terms, but not always in proportion to the total number of articles in respective categories. Conclusions: Although the number of publications has increased over time, the Japanese share of published studies in RCT, CS, CCS and MA categories were lower compared with the average figures for the rest of the world.