Marius Comper

Health · 3,181 localities

Atlas of Medical Deserts

In 2024, Romania had 34.0 doctors per 10,000 residents, up from 21.2 in 2004 — a rise of nearly two-thirds in twenty years. The national figure hides a vast distance between places. Towns have 56.4 doctors per 10,000 residents; communes have 5.8, almost ten times fewer. 315 localities lost their last doctor over this period and have reported none for at least two years, and 566 of the localities that already existed before 2003 have never had a pharmacy or a hospital. The map below shows, for each locality, how many doctors, family doctors, dentists, pharmacies and hospital beds it has per 10,000 residents, today and twenty years ago.

56.4 · 5.8doctors per 10,000 residents in towns and communes, in 2024
315localities with no doctor in 2023 and 2024, despite having had at least one from 2004 onward
27localities whose medical staff went, in twenty years, from entirely public to entirely private
566long-standing localities that have never had a pharmacy or a hospital, since 1995

The map

What medical services exist here?

Search for your locality or click it. The profile shows how many doctors, family doctors, dentists, pharmacies and hospital beds it has per 10,000 residents, how the figure changed from 2003-2004 to 2024, and how its neighbours compare. Pick an indicator and a year, or play through the years.

Loading the map…
0lowhighno recorded population

For each indicator separately, colours split localities with at least 300 residents and at least one such professional into roughly even fifths; localities with none stay in the darkest class, alongside those with no recorded population. On a small screen, pinch to zoom or use the corner buttons.

The country

What grew and what stood still

From 2004 to 2024, doctors per 10,000 residents rose from 21.2 to 34.0: most of the growth came in the last decade, from 24.6 in 2014 to 34.0 in 2024. Not every category grew the same way.

01020301990200020102020202434.0
Doctors per 10,000 residents, nationwide, 1990–2024. The dotted line marks 2004, the starting point of the twenty-year comparison used throughout this page.

Dentists grew faster than doctors overall: from 4.4 per 10,000 residents in 2004 to 9.7 in 2024, more than double. Pharmacies followed a similar path, from 2.1 to 3.9 per 10,000. Family doctors tell a different story: from 5.1 per 10,000 in 2004, they rose to a peak of 6.5 in 2011, then fell to 5.4 in 2018 and reached only 5.8 by 2024 — less than in 2011, thirteen years earlier. The family practice in every commune, the layer of care designed precisely to cover the territory evenly, is the only one that hasn't just failed to keep pace with the rest of the system, but hasn't even recovered its own peak.

0246810family doctors 5.8dentists 9.7pharmacies 3.91995200020052010201520202024
Family doctors, dentists and pharmacies per 10,000 residents, nationwide, from each indicator's first year with data to 2024.

Hospital beds for continuous care have stayed nearly unchanged: 62.8 per 10,000 residents in 2004, a low of 58.6 in 2014, 62.6 in 2024. The hospital system hasn't expanded in twenty years; there has only been a redistribution between units, visible below in the hub localities.

Towns already had a much higher rate than communes in 2004, and the gap hasn't narrowed. In 2024, towns have 56.4 doctors per 10,000 residents, communes 5.8; for dentists, 15.5 against 2.4, more than sixfold. Family doctors are almost the only category that isn't concentrated in towns to the same degree: 7.1 in towns, 4.1 in communes, under double. That is exactly the layer that should reach the territory evenly — and it's also the one that has grown the least.

The counties

From 11 to 86

The doctor rate for a county, computed from the sum of its localities' doctors and residents, runs from 11.1 per 10,000 residents in Vaslui to 85.8 in Bucharest: nearly eightfold. The bottom of the list is almost entirely counties without a large university city with an emergency hospital: Vaslui, Dâmbovița, Călărași, Giurgiu, Ialomița. The top corresponds to counties whose seats are cities with clinical hospitals and medical schools: Timiș (Timișoara), Cluj (Cluj-Napoca), Dolj (Craiova), Mureș (Târgu Mureș), Iași and Bucharest.

Romania's counties by doctors per 10,000 residents in 2024, with each county's population
CountyDoctors per 10,000Residents
Vaslui11.1509,726
Dâmbovița12.1504,683
Călărași12.2294,334
Giurgiu12.7258,527
Ialomița13.6275,751
Teleorman13.8342,258
Suceava13.8764,726
Tulcea14.1222,505
Vrancea14.2371,065
Buzău14.7440,313
Botoșani14.9457,997
Neamț15.3547,845
Bistrița-Năsăud15.8323,634
Bacău15.8717,335
Satu Mare16.4378,629
Prahova17.4757,213
Caraș-Severin17.5300,931
Brăila17.8321,605
Covasna18.7221,294
Olt19.1410,559
Sălaj19.2238,045
Maramureș19.5509,644
Mehedinți19.7266,504
Alba19.7363,660
Harghita19.8324,082
Galați20.4625,067
Vâlcea21.2384,967
Gorj22.0342,226
Ilfov23.8513,905
Argeș25.8610,907
Hunedoara27.8431,005
Arad30.0460,634
Sibiu34.1467,877
Brașov34.9641,572
Constanța38.0748,860
Bihor43.1609,430
Iași50.8993,235
Mureș54.1578,574
Dolj56.2667,708
Cluj60.7744,031
Timiș79.2762,776
București85.82,143,578

A county's rate is the sum of its localities' doctors divided by the sum of their residents; the municipality of Bucharest is counted as a single county, as INS does.

Twenty years

The localities that lost their last doctor

315 localities had at least one doctor in some year between 2004 and 2022, but report none in either 2023 or 2024. This isn't a one-year blip: in each of them, INS recorded zero doctors, two years running, the most that an annual series can show about a practice that closed, or a doctor who retired or moved and was never replaced. The most populated of them are communes of several thousand residents, not tiny villages where one person leaving could be statistical noise.

  1. Suceava20198,173 residents
  2. Botoșani20225,836 residents
  3. Prahova20205,481 residents
  4. Covasna20204,980 residents
  5. Suceava20204,719 residents
  6. Bacău20224,128 residents
  7. Satu Mare20153,993 residents
  8. Suceava20193,969 residents
  9. Vaslui20213,916 residents
  10. Maramureș20213,752 residents

The year shown is the last one in which the locality reported at least one doctor. Click a name and the map jumps there.

The sector

From public to private, without disappearing

In 64 localities with at least six doctors in 2004, the share employed by the state fell by at least 40 percentage points by 2024, while the total number of doctors stayed at least at 70% of its level from twenty years earlier — so this isn't simply a shrinking workforce, but a genuine change of employer. In 27 of them, medical staff flipped completely: from 100% public employees to 100% private.

  1. Brașov100%→0%7→5 doctors
  2. Constanța100%→0%6→8 doctors
  3. Mureș100%→0%6→7 doctors
  4. Mureș100%→0%10→7 doctors
  5. Prahova100%→0%6→6 doctors
  6. Prahova100%→0%7→7 doctors
  7. Timiș100%→0%6→5 doctors
  8. Arad90%→0%10→8 doctors
  9. Arad88%→0%17→12 doctors
  10. Dolj88%→0%8→6 doctors

The percentages show the share of doctors employed by the state in 2004 and in 2024; the numbers show the total doctor count in the same years. Click a name and the map jumps there.

The hubs

The localities that serve more than themselves

18 localities with under 4,000 residents have their own rate of at least 25 doctors per 10,000, inflated by a hospital or medical centre that draws patients from the wider area. Add the neighbours’ doctors and population to the numerator and the denominator alike, and the rate more than halves in every single case. INS counts staff at the institution's registered address, not by where patients come from, so a small commune can look like a medical hub only because it hosts a building that serves a much wider catchment.

  1. Argeș68.9→11.81,597 residents
  2. Cluj62.2→8.81,287 residents
  3. Buzău60.4→11.22,154 residents
  4. Buzău57.6→9.13,472 residents
  5. Argeș65.7→18.03,197 residents
  6. Bihor60.1→16.81,995 residents
  7. Timiș60.8→19.03,786 residents
  8. Călărași62.1→24.92,416 residents

The first figure is the locality's own doctor rate per 10,000 residents in 2024; the second is the rate computed over the locality's population plus its direct neighbours, with their doctors added to the numerator. Click a name and the map jumps there.

Nothing at all

566 localities with no pharmacy and no hospital

Among the localities that already existed before the 2003–2006 administrative reorganisation, and so had thirty years in which a pharmacy or a hospital could have appeared, 566 have never reported one, from 1995 to 2024. They aren't necessarily tiny villages: several of them have close to 5,000 residents.

  1. Galați5,975residents
  2. Botoșani5,322residents
  3. Botoșani5,278residents
  4. Prahova4,832residents
  5. Covasna4,781residents
  6. Botoșani4,723residents
  7. Botoșani4,655residents
  8. Olt4,486residents
  9. Vaslui4,427residents
  10. Alba4,287residents

Most of the other long-standing localities have had a pharmacy or a hospital at some point; communes created after 2003 are excluded from this list so they aren't mistaken for a genuine desert, since a new commune can look "pharmacy-free" simply because the pharmacy stayed registered under the parent commune it split from. Click a name and the map jumps there.

Method and sources

Where the numbers come from

All data comes from the TEMPO-Online database of Romania's National Institute of Statistics (INS), downloaded in full, for all 3,181 localities, on the night of 8-9 September 2026. Medical staff — doctors, family doctors as a subgroup of them, and dentists, by form of ownership — comes from SAN104B, 1990–2024. Pharmacies and hospitals, as units, come from SAN101B, 1995–2025. Hospital beds for continuous care come from SAN102C, 1990–2025. Each locality's population, used as the denominator, comes from POP107D, 1992–2026. The sum of each county's localities was checked against the county's own row, for every table and every year.

Across all three SAN matrices, a missing year in a locality's series means zero for that year, not a data gap: empirical spot checks show localities with doctors in one year, zero for several years running, then doctors again — exactly the pattern of a practice closing and reopening. Family doctors have only been recorded separately since 2003, and dentists since 1993; before those years the category simply wasn't tracked, it wasn't zero, so it doesn't appear on the charts before its start year. A locality's rate is its count of people or units divided by its population in the same year, multiplied by 10,000.

The twenty-year comparison uses 2004 and 2024. Localities created after the 2003–2006 reorganisation only carry a population from their founding year onward; to avoid mistaking a new commune, with no staff inherited from its parent commune, for a genuine desert, the "lost its last doctor" list requires a doctor recorded from 2004 onward, and the no-pharmacy-or-hospital list only includes localities with a recorded population already before 1994. The threshold of six doctors in 2004, used for the public-to-private localities, excludes cases where a single doctor changing employment status would swing the percentage by 100 points. These hub localities, as we call them above, require an own rate of at least 25 doctors per 10,000 and a population under 4,000; neighbours are localities sharing a border, computed from the geometry of the administrative units, and the wider rate is computed from the combined doctors of the locality and its direct neighbours, divided by their combined population.

The outline of the 3,181 administrative units comes from the "UAT, Romania (polygon)" dataset published by the geo-spatial.org community from data of the National Agency for Cadastre and Land Registration, the 26 March 2025 version, under a CC BY-SA 4.0 licence. Bucharest's six sectors were merged, since INS publishes data only for the municipality. The geometry was simplified for display and has no cadastral value. Matching data to outlines used each locality's SIRUTA code and succeeded for all of them.

The figures are descriptive. They show how many professionals and units are registered in a locality, not who treats its residents: someone may see a doctor in the neighbouring town, and a doctor may be registered in a locality without living there. The explanations in the text about commuting, urban concentration and substitution of the public sector are the most plausible ones, supported by the shape of the data, but they aren't measured directly by it.

↑ Top