South Korea’s 490 Regional Doctor Seats Begin a Longer Route to Care

The first cohort will not reach practice until the 2030s. In South Gyeongsang, admissions, mandatory service, specialist training and patient referral already follow different maps.

South Korea has begun selecting the first 490 students for a new system designed to move more of the future physician workforce into regional medicine. Of the places available for 2027, 359 are tied to defined care areas and 131 to broader regional pools. Students receive support for tuition and other education costs and, after obtaining a medical license, must work for 10 years in the service region attached to their admission. Most participating medical schools began their first regional-doctor admissions in September.

The admissions number arrives long before the medical care. Students entering in 2027 will spend six years in medical school before graduating around 2033, and those who pursue specialist careers will continue through residency before reaching independent specialist practice. Hospitals trying to cover emergency shifts, maintain maternity services or recruit physicians to smaller communities have to work with a different pool of doctors through the rest of this decade. The government’s own workforce plan places the additional doctors produced by the 2027–31 expansion in the period from 2033 onward.

Regional medical capacity is produced through more than admission. Physicians choose specialties, residency programs determine where those skills can be acquired, hospitals need enough personnel and infrastructure to use them, and patients move between institutions when local facilities cannot provide the treatment they need. South Gyeongsang offers an early view of those layers because the first regional-doctor allocations can already be compared with a medical system in which specialist training is concentrated in a few hospitals and patients regularly cross the boundaries used for workforce planning.

The First Cohort Enters a Disrupted Pipeline

South Korea held its annual medical-school quota at 3,058 for nearly two decades before sharply increasing it for the 2025 academic year. The official quota rose to 5,058, while the number ultimately recruited was adjusted to 4,567; admissions then returned to 3,058 for 2026. The next expansion is smaller and phased, taking the 2027 total to 3,548 by adding 490 places above the old baseline. Every additional place at the existing schools is being filled through the regional-doctor track.

The intervening years changed more than the size of an admissions class. The confrontation over the 2024 expansion was followed by prolonged disruption in medical education and residency training, leaving universities and hospitals to absorb returning students while another expansion begins. The Korean Institute of Medical Education and Evaluation placed 30 medical schools under major-change review because their enlarged 2025 intake could materially affect medical education. Twenty-six schools that retained accreditation after the second-year review entered a third year of assessment in 2026.

A separate workforce problem has already reached communities that depend on public doctors. Only 98 physicians newly entered the medical public health doctor program in 2026 while 450 completed service, cutting the total number of medical public health doctors from 945 to 593 in one year. The Ministry of Health and Welfare linked the unusually steep decline partly to the medical-school and residency disruptions of 2024 and 2025, although the public health doctor workforce had been shrinking for years before the dispute, from 2,116 in 2017 to 1,209 in 2024.

The regional-doctor program therefore begins on two different timelines. Current hospitals and communities have to maintain services with the physicians available through the rest of the 2020s, while the first new cohort moves through medical school and, for many, specialist training during the following decade. The 490 seats can reshape the second workforce without supplying the first.

The New Seats Come With a Geography

The regional-doctor system moves geographic policy to the beginning of a medical career. A national quota determines how many students enter medical school but does not determine where they eventually practice. The new track connects admission to a defined service region, finances much of the student’s medical education and places a 10-year regional service condition on the physician’s license after graduation. Central and regional support programs are also intended to provide education, career development and settlement assistance.

Two geographic scales operate inside the first cohort. The government assigned 359 places to care-area recruitment and 131 to broader regional recruitment, allowing some students to enter through relatively narrow local boundaries while others qualify through a larger regional pool. Those categories help determine who may enter the program and where the resulting service obligation is attached. Neither should be read as a list of 490 predetermined hospital jobs.

The hospital geography remains partly unfinished. Regulations define categories of institutions in which mandatory service may be performed, but the first specific list of eligible service institutions does not have to be published until Dec. 31, 2029. Students entering medical school in 2027 can therefore know the duration and broad geography of their obligation several years before the institutional map through which that obligation will operate is complete.

Specialty follows another set of rules. South Korea recognizes 26 specialist fields, while the regional-doctor program gives more favorable service credit to training in nine fields when residency takes place within the mandatory region: internal medicine, neurology, surgery, neurosurgery, cardiovascular-thoracic surgery, obstetrics and gynecology, pediatrics, emergency medicine and family medicine. Regional doctors are not assigned to one of those fields according to the shortage in the place attached to their medical-school seat. The system alters the incentives surrounding a later career decision rather than making the decision at admission.

A future physician can consequently be attached to a region years before the physician’s eventual specialty, training hospital and workplace are known. Areas that need specialists may not contain the residency programs capable of producing them, and the rules allow training outside the original service area when appropriate local training is unavailable. Geographic assignment begins early, while much of the clinical pathway remains dependent on choices and institutions that come later.

One National Shortage Becomes Several Local Problems

National workforce planning starts from an aggregate calculation. For the 2027–31 decision, the government selected a model projecting a shortage of 4,724 physicians in 2037 and then accounted for 600 doctors expected from planned public and new regional medical schools. Education conditions and university-level expansion limits were applied before the first increase was set at 490, with another workforce projection scheduled for 2029.

A national number cannot identify the kind of medical service that will be missing. An unfilled pediatric residency position, too few obstetricians to maintain a maternity service, inadequate emergency coverage and a rural community without nearby primary care are all physician-workforce problems, but an additional doctor does not have the same effect in each setting. Hospital organization introduces another layer because specialist care depends on colleagues, nurses, operating rooms, diagnostics, intensive-care capacity and referral relationships as well as the individual physician.

South Gyeongsang’s current residency market already shows the difference between overall recruitment and specialty recruitment. Provincial officials told the assembly in July that the overall resident fill rate at the province’s training institutions was 66.6 percent, while the combined rate in pediatrics, obstetrics and gynecology, emergency medicine and cardiovascular-thoracic surgery was only about 30 percent. The provincial government is supporting residents in those fields because the narrower specialty shortage is considerably more severe than the overall rate indicates.

The national planning process has also begun moving below the aggregate figure. A physician-workforce projection committee established in 2025 is composed of 15 experts, including eight nominated by health-care provider groups, four by consumer groups and three by academic or research organizations. The committee examines models, assumptions and variables behind physician supply and demand rather than treating the medical-school quota as a number generated automatically from a single forecast.

Future physician need will remain partly uncertain even under a formal projection process. Population aging, medical utilization, physician working patterns, technology and the organization of care can all alter demand, while education capacity limits how quickly additional students can be absorbed. A national shortage estimate can establish the scale of a workforce problem without specifying where each additional doctor should train or what medical service each one should provide.

South Gyeongsang Shows Where the Maps Diverge

Six medical schools in the Busan-Ulsan-South Gyeongsang admissions bloc receive 97 of the first regional-doctor seats. The allocation spans Pusan National University, Gyeongsang National University, Dong-A University, Inje University, Kosin University and the University of Ulsan. The 97 describe where medical-school places have been assigned, not 97 doctors already distributed across the three jurisdictions. The Ministry of Education’s regional-doctor admissions guide provides the university and care-area allocation framework used for the first cohort.

Seventy-one of those seats are directly attached to five care areas in South Gyeongsang: 22 to Changwon, 13 to Jinju, nine to Tongyeong, 21 to Gimhae and six to Geochang. The remaining 26 use the broader regional recruitment route. Calculated from the published allocations, the directly assigned seats account for 73.2 percent of the six schools’ 97-seat total. The calculation describes admissions geography rather than eventual hospital placement.

Recruitment geography and service geography are not necessarily the same. Students may qualify through a broader Busan-Ulsan-South Gyeongsang pool even when the mandatory service attached to particular seats points toward South Gyeongsang. A medical school located in Busan or Ulsan should therefore not be treated automatically as producing a regional doctor whose obligation will be served in that city. The policy separates where a student studies, where the student was eligible to apply and where the physician will later have to work.

Training capacity draws another map. Immediately before the resident walkout in February 2024, South Gyeongsang had 479 residents across 10 training hospitals, but 447 were concentrated in four institutions: Pusan National University Yangsan Hospital, Gyeongsang National University Hospital in Jinju, Samsung Changwon Hospital and Gyeongsang National University Changwon Hospital. More than 93 percent of the province’s residents were therefore working in those four hospitals before the subsequent disruption. The figures are a historical baseline rather than a current 2026 headcount, but they show how concentrated the specialist-training structure already was.

Tongyeong and Geochang together receive 15 of the 71 directly assigned seats, or 21.1 percent. Neither care area contains one of the four hospitals that accounted for almost all of the province’s residents in that February 2024 snapshot. Future doctors attached to those areas may train through smaller institutions, new regional networks or programs elsewhere before returning. Their route from a medical-school allocation to specialist practice will nevertheless differ from the route available around the larger training centers in Yangsan, Changwon and Jinju.

Gimhae exposes another kind of divergence. The regional-doctor framework groups Gimhae with Miryang and Yangsan inside a wider policy geography, but existing specialized-care patterns differ sharply within that area. National Health Insurance Service data for 2023, reported this month, put the share of severe pediatric inpatient treatment obtained within the patient’s own intermediate care area at 81.0 percent in Yangsan and 6.8 percent in Gimhae. The same data show that movement from Gimhae toward Yangsan contributes to the difference.

Those percentages do not rank the overall quality of the two local health systems. They cover severe pediatric inpatient care, and the health-insurance geography does not perfectly reproduce the regional-doctor admissions boundary. Transfers to a stronger pediatric center can also be appropriate care rather than evidence that every severe case should remain close to the patient’s home. The narrower finding is that municipalities grouped inside one workforce policy area can already depend on very different medical networks for the same category of treatment.

Regional health systems are not improved by eliminating every journey across a boundary. Complex surgery, trauma, high-risk maternity care and some severe pediatric conditions are often safer when concentrated in institutions with sufficient case volume and supporting specialties. Smaller hospitals can improve access by maintaining services that can reasonably be sustained locally and by transferring other patients quickly to hospitals able to provide the care they need.

Regional-doctor performance should therefore not be reduced to medical self-sufficiency within each administrative area. The more useful question is whether locally sustainable services become more available near residents while referral networks improve for treatment that should remain concentrated. The first seat allocation cannot yet reveal either result.

Specialty Becomes a Second Allocation

The first allocation takes place when the state distributes medical-school seats. Another occurs years later when those students choose specialties and enter residency programs. South Gyeongsang’s 66.6 percent overall residency fill rate alongside a roughly 30 percent rate in four difficult fields shows how strongly the second distribution can reshape the effect of the first.

The regional-doctor system attempts to influence that choice through the 10-year service calculation rather than direct specialty assignment. Training in selected fields can consume more of the mandatory-service period, changing the relative cost of one career path compared with another. The mechanism may draw physicians toward fields facing shortages, but the actual distribution will not be known until regional-doctor students begin selecting residency programs.

Training availability then constrains the options. An area can need a specialist without containing the hospital capable of producing one, forcing the physician to acquire skills elsewhere before returning to a regional post. The concentration visible in South Gyeongsang’s pre-2024 residency data means that some care areas begin the program with a much thinner specialist-training base than others.

Returning with a specialist certificate still does not complete the conversion. Hospitals need funded positions, enough colleagues to maintain schedules and the complementary staff and facilities required for the service. An obstetrician alone cannot sustain high-risk maternity care, and an emergency physician alone cannot provide definitive treatment for every surgical or intensive-care case that reaches an emergency department.

The same admissions total can therefore produce several different medical systems. Regional doctors may enter specialties that match the hardest local shortages or fields with less severe shortages, strengthen smaller hospitals or join existing regional centers, and work in institutions that either expand services or remain dependent on the same referral networks. Those outcomes all begin with a medical-school seat, but the seat itself cannot distinguish among them.

Workforce Planning Extends Beyond the Quota

South Korea’s medical-school quota carried an unusual amount of the debate over physician supply because it remained fixed at 3,058 for so long. The 2024 attempt to add 2,000 places intensified the dispute around one national number even though the production of physicians was already divided among medical schools, accreditation bodies, residency programs, hospitals and individual career decisions. The later establishment of a standing projection committee makes the assumptions behind future workforce estimates more explicit without placing every downstream decision under central control.

Formal authority over admissions differs from control over what follows. Government can create seats and attach regional service conditions, but universities must educate the students, accreditation bodies evaluate their programs, residency hospitals produce specialists and regional hospitals assemble the teams in which those specialists practice. Professional organizations can contest or influence policy, while individual physicians still make career choices within the rules that government and training institutions establish.

International experience shows why those stages should be evaluated separately. Japan has long used regional medical-school quotas, prefectural scholarships and service commitments to influence physician distribution. A nationwide study found that 92.2 percent of graduates who entered through both a regional quota and scholarship remained within their contractual workforce about 3.25 years after graduation. The figure describes compliance with the contract during the observed period, not permanent retention after the obligation ended.

Australia has also used bonded medical places to direct doctors toward areas of workforce shortage. A 2024 government review found limited public evidence that the bonded scheme and its geographic classifications were achieving the intended distribution on their own, and described the certainty of evidence for compulsory-service agreements alone as low. The review emphasized a broader combination of rural background, clinical education, training experience and other workforce policies.

Neither comparison predicts what South Korea’s 10-year obligation will produce. They separate outcomes that can otherwise be compressed into one question about whether a regional-doctor program works. Contractual placement, specialty choice, hospital capacity and retention after mandatory service measure different stages of the workforce pathway.

South Korea’s program reaches farther into that pathway than a national quota increase alone. The state is deciding not only how many additional students enter medical school but also where part of that future workforce must serve and which training choices receive stronger incentives. The route remains only partly governable from the center because each additional stage introduces institutions, professional choices and local health-system conditions that an admissions decision cannot settle.

The Evidence Arrives Before the Doctors

The first regional-doctor students will not graduate until around 2033, but the program does not have to remain unevaluable until then. Admissions results will show who enters through the narrower care-area and broader regional tracks, while continuing medical-school reviews will show how institutions absorb the expanded classes. Decisions on service institutions and postgraduate training will provide evidence about the next stages long before the first cohort completes a decade of mandatory work.

A major checkpoint arrives in 2029. The government must publish the first specific list of institutions eligible for mandatory regional service by Dec. 31 of that year, and another national physician-workforce projection is scheduled before the following quota cycle. The first cohort will still be in medical school when both the estimated size of the future workforce and the institutional geography of its eventual service can be reconsidered.

Residency will produce another distribution. The relevant numbers will no longer be 490 nationally or 71 across five South Gyeongsang care areas, but how many regional doctors enter particular specialties, where they can train and how often a lack of local training capacity sends them elsewhere. The incentives written into the program can then be compared with actual career choices rather than with the design of the policy on paper.

Hospital employment will reveal another layer. A returning specialist can strengthen a smaller institution, join a major regional center or take another eligible post within the service geography. Each may satisfy the legal obligation while producing a different change in access for patients, depending on the clinical services the hospital can sustain.

Patient movement belongs later in the evaluation because workforce distribution is not identical to the geography of treatment. Less travel for services that can safely and sustainably be provided closer to home can indicate new local capacity, while faster transfer to a capable regional center can indicate improvement in services that should remain concentrated. Referral is not the opposite of regional care; for many complex conditions, it is one of the ways regional care works.

South Gyeongsang already contains the components needed to follow that conversion. Five care areas have assigned seats, historical residency data identify where specialist training has been concentrated, and current patient-use evidence shows that places inside the same broad policy geography can rely on very different medical networks. The 2029 service-institution list will add another layer, followed later by specialty choices, residency locations and actual hospital placements.

The regional-doctor system connects medical-school admission to geography, finances students through their education and attaches a decade of regional service to the physicians who emerge. Those mechanisms create a more direct route toward regional workforce distribution than a national quota increase alone, but hospitals and patients encounter only what survives the stages that follow. A regional-doctor seat becomes regional medical capacity when specialty, training, employment and access connect.