According to the Sophia Globe (h/t Pileta), the oldest ever city in Europe would have been located at Provadia (Varna province, Bulgaria).
The city is defined as such by medieval standards because it is a settlement surrounded by a wall, a very thick and solid one. However only a few hundred people (300-350) lived inside them, always according to Prof. Vassil Nikolov.
Nikolov attributes the city an age of 4700-4200 BCE, belonging to the context of
Varna culture and related to the relatively well known
Varna necropolis, probably the first princely burial of the continent.
However such an early age places the Provadia-Solnitsata city almost two millennia before any other known European city (Greek islands, Southern Iberia) and even 1500 years before
Troy I, traditionally considered a major influence in the urbanization and Bronze Age of Europe.
Not even Egypt’s civilization is nearly as old; in all the wider region only some fortified cities from the Levant and Mesopotamia are as old and, excluding
Jericho (c. 6800 BCE), none is significantly older. Even the oldest known Sumerian city,
Eridu, was built only a few centuries before this Bulgarian city.
For more details, read
the abstract (PDF) of Nikolov’s study (to be published?, where?)
Looking for more information on the matter I also stumbled upon
this webpage by Dr. L. Nikolova, which discusses the Provadia settlement as a salt-exporting site, something that also seems to be the leit motif of Nikolov’s abstract.
Discussion
I am not really surprised by these findings because it has been known for some time that the area of Bulgaria was a very old center of civilization, comparable in age to Egypt (or, as it seems to be the case now, even quite older), probably evolving state and aristocracy structures first of all in Europe. It was probably the wealth of this realm and its successors which baited the Indoeuropean nomads into invasion later on (economic relations with the Volga are attested as early as this time).
The only thing that really puzzles me is the age because when I first learned on these matters the time-frame appeared to be a thousand years more recent, more in agreement with other urbanizing and social-complexity developments elsewhere in Europe (Aegean, Iberia), Western Anatolia (Troy) and also Africa (Egypt). But well… I have to accept the dates estimated by the researchers, I just wish I knew a bit more about how these have been produced.
Update (Oct 10): photos and considerations on burial styles
Terrae Antiquae[es] has now a very extensive photo-gallery, of which I borrowed the following:
I must say that unlike the Varna necropolis burial which I initially used to illustrate the news, the burials shown here appear to be of the classical Neolithic flexed style (the lower part may be missing but the size of the pit and lateral deposition strongly suggest that.
I’m not just inferring from that and another photo but it was actually the burial style of all Balcano-Danubian Neolithic peoples and even some “Danubianized” Indoeuropeans later on: flexed lateral burial, so it is what should be expected in this context.
The extended position is actually typical of Paleolithic Continuity peoples, and the use of ochre is specially documented in Eastern Europe: Dniepr-Don Neolithic and related groups like Pitted Ware and the Early Bronze culture of Ezero also in Bulgaria but thousands of years after these layers.
At the Varna museum site however they do explain that the burials in what was surely a royal or otherwise princely necropolis are of two kinds: (a) 99 burials (mostly men) are in extended position (Paleolithic tradition), while other 67 (many of them women) lay on their right side in flexed position (Neolithic tradition).
This could point (my best guess here) to early penetrations from the steppe (Dniepr-Don culture), comparable maybe to those we see around the Baltic (Pitted Ware), all however anticipating and maybe preparing the way for what is probably the true Indoeuropean penetration of the Kurgan wave of cultures, which in most cases would return to the flexed burials… now in kurgan (tumulus).
However in Bulgaria the period of Kurgan invasions culminates in the formation of Ezero culture, which is the only one retaining extended burial with ochre, by that time already vanishing in the steppe and the Baltic.
For what they were... we are 
Racer
October 9, 2012 at 5:07 am
Any information yet or independent confirmation on the age of the town?
Maju
October 9, 2012 at 6:58 am
No. I just posted that hours ago. It looks credible but I would indeed want more info – as you probably do as well.
Maju
October 10, 2012 at 4:31 pm
Update: some photos and mini-discussion on the burial styles here and in Varna necropolis.
rouge77
October 10, 2012 at 9:08 pm
It looks like a fortified village to me. Impressive one, but not a town. Mere walls don't make a town.
Maju
October 10, 2012 at 11:12 pm
That's a matter of opinion. Today we think of a city or town because of size but in the Middle Ages (and often earlier) it was their walls which defined a city. There were exceptions but those were usually in islands (Crete notably). According to N.J. Pounds, Europe was in the Middle Ages full of thousands of small towns, defined as those under 2000 inhabitants. He mentions Rheinfelden (Switzerland, 15 km south of Basel) as example. Characteristics: walls, charter, 220 homes (c. 1000 inhabitants), some 10 smiths, 20-30 leatherworkers, 20-30 weavers, masons, carpenters. Estimating that 2/3 of adult men were artisans, the rest would be agriculturists, as would be part-time many artisans.So besides walls, what characterizes a town? Artisans and maybe services. And I think that we can be pretty sure that in Chalcolithic Bulgaria there were many artisans: coppersmiths, goldsmiths and many others less well attested but no doubt part of the social matrix. It's probable that these small towns also held regular markets for the area, maybe weekly for the agricultural products. … But then, on second thought, it's possible I guess that it was a castle or acropolis and not the full settlement. That's the impression I get from the photos, specially those showing human scale. If so we'd be before something else (maybe bigger, maybe not) but more a castle than a true town. Would it be a town with all its industry and services, it'd be very compact indeed.
DDeden
October 11, 2012 at 5:03 pm
http://www.iianthropology.org/saltprehieurasia.htmlThe locale and salt industry are critical(Koros-Cris|Horus-Christ|Amaterasu|Moses myth origin linked to the salt trade via danube/jordanubya-sudanubya/Llubjana(S-lovenia)-Lebanon/Albania)http://the-arc-ddeden.blogspot.com/2012/03/danubya.html
Maju
October 11, 2012 at 6:19 pm
Please avoid that pseudolinguistic nonsense. Köros and Cris(h) are two different sites, the names are sometimes used together to honor both th Hungarian and the Romanian side of the border. They are anyhow just variants of Starcevo, a Serbia-centered culture.
Raboti Makedontskite
October 12, 2012 at 5:47 pm
Sesklo in Greece is a site that showes signs of an "urban" settlement and it is older than this Bulgarian find, even if it would be to accept as accurate its already established age by Nikolov.
Maju
October 12, 2012 at 7:47 pm
Maybe true. While the initial Sesklo is not urban at all, the late Sesklo (Sesklo B) shows a wall and apparent social stratification (different kinds of pottery) as could correspond to an urban or quasi-urban kind of settlement. However this Sesklo B is almost contemporary with the Bulgarian find, being destroyed by fire (as many other settlements) c. 4400 BCE, roughly the same date given to Provadia-Solnisata. If anything it'd be just slightly older. Ref. http://www.ime.gr/chronos/01/en/nl/mn/sesklo.html
andrew
October 15, 2012 at 8:48 am
I'm curious about the source of the estimate of household size to reach the 1000/220 homes estimate.I've been reading Albion's Seed, which is mostly a history of 18th century early colonial North America, and on of the interesting statistical tidbits it provides is that average household size was robustly 7 people across the Americas even though there were very different numbers of children of the householding couple in different regions (ranging from a bit more than two all of the way to five).Given that 18th century colonial America (which was 90%+ farmers with few large cities in most regions) had demographics (e.g. life expectency and child mortality rates) much closer to Neolithic era farming life than any other reliable data set at my disposal, I wondered is average household size might be closer to seven than four in the early Neolithic.Obviously, there are all sorts of reasons that the facts could be different, but I don't really know what data we do have on household size in the Neolithic or for that matter, on any pre-Roman period where statistical record keeping improves (although I suppose that some estimate of household size could be teased out of early copper and Bronze age ration accounts, some of the very earliest known writing).
Maju
October 15, 2012 at 10:25 am
That's N.J.G. Pounds himself but 5 people per household it is a standard for "translation" of the Medieval homes' figures into approx. inhabitants in other authors also. Now, that you or Seed think otherwise, I'm fine with it, specially if well documented. Notice anyhow that there were also households of bachelor people, couples without children, older people whose children had mostly established themselves apart… and all them counted as fiscal homes. But if you think 7 is a better figure, I don't feel able to discuss it.