BACKGROUND The aim of this study was to evaluate the academic performance and research involvement of students coming from urban and rural backgrounds. To our knowledge, this is the first such study of undergraduate medical students in Croatia. METHODS We surveyed the final (sixth) year medical students from Medical School, University of Zagreb, Croatia. Students were surveyed during the academic year 2004/2005, several months prior to their graduation. We analysed students' academic performance (grade point average, which is the most important academic success indicator in Croatia, and the number of failed study years) and research involvement (involvement in research projects and the possible subsequent publication of scientific articles). Additionally, we investigated the extent of extracurricular activities, and students' workplace preferences. Data were analysed using chi2 test, Fisher's exact test, and Kruskal-Wallis test, due to non-normal data distribution. RESULTS A total of 204 students (out of 240 enrolled, with a response rate of 85%) were surveyed, and divided into three groups: (1) those coming from the highly urbanised background (capital and other highly urbanised areas; n = 100, 49%); (2) mid-urban (towns; n = 75, 37%); and (3) rural and remote backgrounds (n = 29, 14%). There was no indication of gender gap or age difference among the three groups. However, significant differences were recorded in most academic and research indicators. Students from highly urbanised backgrounds reported the best grade point average (p = 0.022). Students from rural and remote backgrounds most commonly reported a study year failure (p = 0.032), with 17 (59%) cases, compared with 32 (32%) and 31 (42%) cases in high and mid-urban ones, respectively. Rural and remote students were also the most likely to experience multiple years failure (p = 0.030), were the least often involved in research projects (p = 0.002), and reported the least interest in career supplemented with scientific research (p = 0.015). Finally, students from rural and remote backgrounds were the least commonly involved in extracurricular activities, such as membership in student organizations (p = 0.005), or various hobbies (p = 0.009). However, 12 (41%) of rural and remote students were likely to seek employment in rural and remote areas, compared with a much lower percent among high and mid-urban students, with 16 (16%) and 16 (21%) students, respectively (p = 0.014). CONCLUSIONS This study exhibited significant differences in the academic performance and research involvement of students coming from urban and rural backgrounds. Students from highly urbanised background reported the best academic indicators, while students from rural and remote backgrounds reported the poorest. Implementation of a rural students' support program should be strategically planned and carried out, because students from rural backgrounds in more than half the cases experienced a study year failure. Because students from rural and remote backgrounds were the most likely to prefer employment in rural and remote areas, they present a valuable resource for reduction of disparities in health workforce geographical distribution.