Methods

Participants

The study was approved by an institutional review board (application no. 144192) on the basis that all data were openly available. An observational design, using data from online sources (https://worldathletics.org), was used to rank and evaluate the achievements of athletes competing in running events held as part of the Olympic program. Athletes were included if they had won at least one medal at a major global championship [12] or set a WR ratified by World Athletics [13] as of 1 March 2026. World Best marathon times [13] from before 2004 were considered WRs provided they were accepted by the International Association of Athletics Federations (IAAF), since renamed World Athletics. As the IAAF was founded in 1912 with a purpose of ratifying WRs [14], we excluded WRs set before 1912. Analyzed athletes comprised 1294 men and 824 women who competed in sprint, distance, and hurdling events. These included obsolete events such as the 60 m, 200-m hurdles, various steeplechase races, women’s 80-m hurdles (replaced by 100-m hurdles) and women’s 3000 m (replaced by 5000 m). Performances in team events like relays and the obsolete 3000-m team race were excluded. Those athletes who won medals in the three editions of the Olympic men’s individual cross-country (1912–1924) were incorporated in the tally for the World Cross Country Championships (WXC). Athletes’ nationalities were indicated using their World Athletics member federation codes [13], thereby showing the names of federations that existed at the time. Data were collated and points calculated using Microsoft Excel; conditional formatting was used to check for duplicate entries, and online sources (https://en.wikipedia.org/wiki/, https://www.olympedia.org/) were used to double-check athlete performances. This was particularly useful for identifying those athletes who won medals under different surnames or for more than one nation.

Data Analysis

Points were awarded for medals on the basis of the value of prize money awarded by World Athletics for the 2025 World Athletics Championships, 2025 World Indoor Championships, 2026 World Cross Country Championships, and World Records (set during the 2025 World Championships) [15–17], regardless of what year the competition was held in. We awarded 1 point per US $1000 of prize money as presented in Table 1. No prize money is awarded in the same way for the Olympic Games, so we decided to use 100 points for a gold medal so that it matched the WR points system and used the same ratios for silver and bronze medals used for the World Championships (Table 1). Although World Athletics did pay US $50,000 to gold medalists in Paris 2024 [18], we did not add any extra points to the default 100 points for Olympic victories. A greater points allocation was awarded to Olympic medals than those in the WC given their less frequent occurrence (every 4 years compared with the biennial WC) and their greater importance to national sporting funding bodies [8, 19, 20]. Overall, the approach taken was a descriptive study using the sum of allocated points to determine rankings.

Championship | Gold | Silver | Bronze | Other
Olympic Games | 100 | 50 | 31 | 
World Athletics Championships | 70 | 35 | 22 | 
World Athletics Indoor Championships | 40 | 20 | 10 | 
World Cross Country Championships | 30 | 15 | 10 | 
World Record |  |  |  | 100
World Record + Olympic gold |  |  |  | + 5
World Record duration |  |  |  | + 2 per year

WRs were obtained only for those running events that have been part of the Olympic outdoor track and marathon program. WRs for the 200 m were included from 1951 onward, as previous WRs could be set on a straight, rather than a curved track [13]. For women, whose first Olympic athletics competition was in 1928 [21], WRs were taken from when those events were recognized by the IAAF. Women’s WRs were recorded for the 80-m hurdles and 3000 m until their replacement by the 100-m hurdles and 5000 m, respectively. Although the first IAAF WC were inaugurated in 1983, the women’s 400-m hurdles and 3000 m from the 1980 World Championships in Athletics were included. Similarly, we incorporated the IAAF World Indoor Games (1985) with the World Athletics Indoor Championships (WIC). We also included the indoor 200-m race in our analysis, which was discontinued after 2004. We included women’s marathon WRs from 1970 onward, as this was the period when women’s marathons became formalized [22], and points were awarded for WRs set in women-only races. Given the truly outstanding nature of setting a WR during an Olympic final, 5 bonus points were awarded to any athlete achieving this feat. To factor in the quality of a WR, a longevity bonus was awarded: for each full year that a WR was held by the same athlete, an additional 2 points were awarded. In those cases where a particular athlete broke their own WR, or held it on multiple occasions (i.e., regained it from another athlete), the cumulative duration of their WR durations was used. Joint WRs were common in the sprint events when hand-timing was used; these were all considered WRs. All points awarded for medals and WRs were based on the official results for those events; athletes whose medals or WRs were subsequently rescinded were not awarded points for those achievements. Athletes who were banned for a period of time for doping violations were included in the sample for any results that were ratified despite the ban.

To aid in establishing the robustness of the model used, a test of sensitivity analysis was conducted. In this process, several different scenarios were run where the contribution of particular achievements were adjusted to identify what changes would occur in the ranking of the top 10 men and women. These scenarios were: (1) an increase in OG points so that they were double the WC points (because the frequency of OG is half that of WC) and WR points adjusted to 140 points to match the new OG gold score; (2) no bonus for setting a WR during an OG gold medal winning race; (3) no bonus for WR duration; (4) no points awarded for cross country (XC) medals; and (5) no points awarded for WIC medals. Although there are innumerable possibilities for altering the points system, these five scenarios were felt sufficient for testing the overall points structure adopted.

To account for historical differences, we defined five distinct racing eras (Fig. 1). These eras overlapped with those identified in a previous historical analysis of Olympic 1500-m performances [23] and are related to the number of athletes competing, the number of championships and events available, the impact of both World Wars, the effect of globalization, and the removal of the amateur code in athletics. Each era was defined by Olympiad years. We grouped the first five OG as the “early era” (1896–1912), which was pre IAAF with only a small number of men representing relatively few nations competing in a nonstandardized range of events. The next five Games (1920–1936) form the “interwar era,” which featured the recognized men’s events that continue today (but included cross-country races) and, from 1928, two running events for women. The Games from 1948 to 1968 were deemed the “post-war amateur era,” where the number of running events for women increased from three to five but when many successful athletes retired early to focus on professional careers [23]. The “early professional” era ran from 1972 to 1992, during which Olympic events for women first included the 1500 m (1972), 400-m hurdles, 3000 m and marathon (1984), 10,000 m (1988), and the 80-m hurdles increased to 100 m (1972). It was during this era that the WC, WXC, and WIC were first held. The final era identified was the current “contemporary era,” starting in 1996, when the women’s 5000 m replaced the 3000 m, the women’s 3000-m steeplechase was introduced (in 2008), and the WXC became a biennial event after 2011. Several recent championships were either postponed or canceled because of the coronavirus disease 2019 (COVID) pandemic [24]. Athletes were allocated to an era on the basis of the first year they won a medal or set a WR, meaning that some spanned more than one era; the total score across eras was used in these instances.

Fig. 1: Five identified racing eras for men and women since the inception of the first modern Olympic Games in 1896

As well as proposing the overall male and female GOATs, and those during each era, we also described the GOAT for event groupings. For the Olympic program, these comprised the sprints (60–400 m), hurdles (80–400 m), middle-distance (800–1500 m), steeplechase (mostly 3000 m but also early variants), long-distance track (5000–10,000 m), and marathon events; we also had categories for cross-country and indoor running (all events). The since-replaced women’s 3000 m was included in the long-distance category. When assessing the GOAT in these categories, only those medals and WRs achieved over the specified distances were included, meaning that events like the 1000 m, mile, and half-marathon were excluded. Because certain events were often “doubled up” by athletes, we compared feasible doubles by pairing race distances by increasing distance, i.e., 100 m and 200 m, 200 m and 400 m, 400 m and 800 m, 800 m and 1500 m, 1500 m and 5000 m, 5000 m and 10,000 m, and 10,000 m and marathon. The number of doubles achieved during the past two eras was calculated and the distances compared for duration, mean running speed, approximate aerobic contribution for men [25–27], and similarity to other championship events.