Conservation Spain

Earliest copy of Mona Lisa found in Prado

Experts say the painting was completed at the same time as Leonardo’s original

A detail of the nearly-conserved Prado copy of the Mona Lisa (Photo: © Museum Nacional del Prado)

A copy of the Mona Lisa has been discovered in the Prado which was painted in Leonardo’s studio—created side by side with the original that now hangs in the Louvre. This sensational find will transform our understanding of the world’s most famous picture.

Conservators at the Prado in Madrid recently made an astonishing discovery, hidden beneath black overpaint. What was assumed to be a replica of the Mona Lisa made after Leonardo’s death had actually been painted by one of his key pupils, working alongside the master. The picture is more than just a studio copy—it changed as Leonardo developed his original composition.

The final traces of overpaint are now being removed by Prado conservators, revealing the fine details of the delicate Tuscan landscape, which mirrors the background of Leonardo’s masterpiece. Darkened varnish is also being painstakingly stripped away from the face of the Mona Lisa, giving a much more vivid impression of her enticing eyes and enigmatic smile.

In the Louvre’s original, which will not be cleaned in the foreseeable future, Lisa’s face is obscured by old, cracked varnish, making her appear almost middle aged. In the Prado copy we see her as she would have looked at the time—as a radiant young woman in her early 20s.

Leonardo da Vinci, and particularly his masterpiece the Mona Lisa, attracts endless sensationalist theories. However, the discovery of the contemporary copy has been accepted by the two key authorities, the Prado and the Louvre.

Uncovering the truth

Until recently, curators at the Prado had no idea of the significance of their copy of the Mona Lisa. There are dozens of surviving replicas from the 16th and 17th centuries. The Madrid version was believed by some specialist to have been painted fairly early, but the absence of the landscape background meant that it aroused little interest (there is no substantive entry on it in the Prado’s collection catalogues).

Although the portrait is finely painted, the dull, black background had a deadening visual effect on the image of the young woman. The sitter is generally believed to represent Lisa Gherardini, the wife of the Florentine cloth merchant Francesco del Giocondo.

The Prado’s painting was until recently assumed to be on oak (rarely used in Italy at the time) and therefore a work by a northern European artist. José Ruiz Manero, the author of a study of Italian art in Spanish collections, concluded that the picture was Flemish.

Last year, the panel was examined and found to be walnut, which was used in Italy (as is poplar, used for the original of the Mona Lisa). In size, it is close to that of the original: the Louvre’s painting is 77cm x 53cm and the Prado’s copy 76cm x 57cm.

In a paper presented two weeks ago at a technical conference at London’s National Gallery, coinciding with its exhibition “Leonardo da Vinci: Painter at the Court of Milan” (until 5 February), conservators revealed that they had discovered that the black background was a later addition. This conference was not covered in the media (for a report, see our February print edition).

A striking photograph was presented at the conference, showing the picture’s condition after 90% of the black overpaint had been removed, leaving just a small section in the upper right. Visually, the landscape transforms the work, bringing the picture to life.

There was an even greater surprise: infrared reflectography images of the Prado replica were compared with those obtained in 2004 from the original of the Mona Lisa in the Louvre. This process enables conservators to peer beneath the surface of the paint, to see underdrawing and changes which evolved in the composition.

The underdrawing of the Madrid replica was similar to that of the Mona Lisa before it was finished. This suggests that the original and the copy were begun at the same time and painted next to each other, as the work evolved.

Identifying the painter

It is quite possible that Leonardo’s assistant met Lisa and may even have been present when she sat for the master. Although no drawings survive, Leonardo probably began by sketching her face and pose. She may also have come to the studio when finishing touches were being applied to the face in the painting.

The Prado's technical specialist Ana González Mozo describes the Madrid replica as “a high quality work”, and in the paper she presented at the London conference, she provided evidence that the picture was done in Leonardo’s studio. The precise date of the original is uncertain, although the Louvre states it was between 1503 and 1506.

Bruno Mottin, the head conservator at the Centre de Recherche et de Restauration des Musées de France, believes that the most likely painter of the Prado copy was one of Leonardo’s two favourite pupils.

Mottin proposes that it was either Andrea Salai, who originally joined Leonardo’s studio in 1490 and probably became his lover, or Francesco Melzi, who joined around 1506. If the Prado replica is eventually attributed to Melzi, it suggests a late date for the original.

What the copy reveals

The Madrid copy of the Mona Lisa is important for what it tells us about Leonardo’s studio practice. The production of a second version, painted alongside the original, is intriguing. It adds credence to Martin Kemp’s theory that Leonardo may also have had a hand in both versions of The Madonna of the Yarnwinder, 1501-07, one owned by the Duke of Buccleuch and the other by a New York private owner (formerly in the Lansdowne collection).

But what is most exciting about the Prado replica is what it reveals about Leonardo’s original. In the Madrid copy there are areas that are better preserved than in the Louvre painting. The replica gives us more detail of the spindles of the chair, the frill on the edge of the fabric on Lisa’s chest and the semi-transparent veil around her left shoulder, arm and elbow.

The Prado's curator Miguel Falomir believes the replica can probably be identified as a portrait listed in the 1666 inventory of Madrid’s Alcazar Palace, although it remains unclear when it first reached the Spanish royal collection.

Coming into the light

Falomir suspects the black overpaint was probably added in the mid-18th century. The reason for this addition is obscure, since the background landscape remained in good condition and Leonardo’s original painting was already very highly regarded. The overpaint may have been added to integrate the copy into an interior with other portraits set against dark backgrounds.

During the past few months, this black covering has been painstakingly stripped away at the Madrid conservation studio, with the final area of dark overpaint due to be removed in the next few days. Later varnish has also been taken away from the rest of the picture, most importantly the face.

The fully conserved replica is expected to be unveiled at the Prado in Madrid in mid-February. It is then due to be loaned to the Louvre in Paris, as a late addition to its exhibition on “Leonardo’s Last Masterpiece: The Sainte Anne” (29 March-25 June). There it will be seen in the same galleries as the original, giving specialists and visitors the first chance to compare the two works. After 500 years, the two versions of the Mona Lisa from Leonardo’s studio will be reunited again.

To see a comparison of the two works side by side, see our online picture gallery

The Louvre's Mona Lisa
More from The Art Newspaper


9 Dec 14
16:54 CET


What other methods apart from "guessing" the time of the painting have the curators used?. The article doesn't state this, as it is very important. Even if it was from the period of time it says it is, I find very odd the eyebrows but overall, the eyelashes, they look so strange, three little lines splattered there almost as if it was forced, it looks funny as in, the whole painting looks beautifully done and then... there goes the amateur looking eyelashes. Why oh why is the only question.

17 Dec 13
16:31 CET


Its a copy, not a bad one, but it has not the serene beauté and facial caracter from the Mona Lisa from the Louvre, and they say their exist more as 8 very fine copy's in the world.

27 Sep 13
18:0 CET


REGARDING MONA LISA'S IDENTITY AND ISABELLA OF ARAGON: A long time ago, Robert Payne identified the Mona Lisa as a portrait of Isabella of Aragon / Isabella d'Aragona (see Payne's book 'Leonardo', 1979). Payne's theory then was repeated by Maike Vogt-Lüerssen (see Vogt-Lüerssen's book 'Wer ist Mona Lisa? Auf der Suche nach ihrer Identität', 2003). To tell the truth, I don't like the said theory, it's completely incongruous. Be that as it may - Payne's authorship is unquestionable.

24 May 13
12:52 CET


An argument can be made for the painting being a portrait of Caterina Sforza ... Magdalena Soest's sensational identification of the "Mona Lisa" as Caterina Sforza was first published in 2002 and bases upon revolutionary discoveries made by the author herself. Let me give a few examples: Leonardo's "Mona Lisa" possesses a perfect physiognomic correspondence with an older(!) portrait, with Lorenzo di Credi's "Caterina Sforza". In the embroidery and the folds of thegarment, in the hairstyle and in the landscape there are emblems of the Sforza (lily, serpent) and the Medici (lily, oval, interwoven rings) represented - the dynasties to which Caterina, through birth and marriage, belonged to and identified with. A Milanese archive document of 1525 (a solicitor's document in a legal battle concerned with the inheritance of Leonardo's student Salai) contains a short description of the "Mona Lisa" and calls the portrait "Quadro dicto la honda. C.": "Portrait of the worthy C."!

5 Oct 12
14:59 CET


Later we will find out that Leonardo did not like the new landscape and that by his own hand, had later blackened it out.

28 Sep 12
15:46 CET


I can not believe that no one can see the masters gift. Study were the light in the painting is coming from. If you are artisticly talented and look at the painting as though you were the artist you would see the secrect of the masters. Follow the light sourse you can do it on all master works. How is the subject lit? Where is the light coming from and remember what did they use for lighting? These works were done by two different artist. One using lamps the other more modern lighting.

20 Aug 12
3:47 CET


It is so hard for me to believe that two artists 1 student and the other a master could paint exactly the same at no fault at all! Unless you are painting by numbers then it's possible to have an identical pieces to the detail, but side by side looking at the model and everything came out the same? It's amazing!

25 Jul 12
18:23 CET


The landscape and style looks Chinese- I agree with Fiorentino Mari above- it resembles nothing of any Tuscan landscape I know. What a truly remarkable work.

12 Apr 12
14:46 CET


This is a very informative article. Good for my class. Thanx.

30 Mar 12
17:5 CET


I own a “copy” of a Murillo´s madonna. Time ago, an art specialist came home, and after taking down the work from the wall and seeing the canvas back and the frame, she told me that the work was not a copy of 18th or 19th century, but 17th century, probably contemporary to the “original”. My work matches the original with outstanding fidelity even in the smallest details, except one thing: the Child, in my work, holds a small crucifix in his hand, and the original does not show this cross. All other details, face´s expressions, fabrics bendings, etc, are absolutely matching the original work. I would like to establish the authorship of the work, because now I strongly suspect that my work is not a simple copy but something else. Can anybody help me?

28 Mar 12
15:47 CET


many thanks for the informative articile. i am about to paint one similar character and was wondering if Mona Lisa or Giaconda was as old as she looks in the Original it is a pity that it has not as yet been cleaned up.

7 Feb 12
14:52 CET


Do we really know for sure which is the "original?"

6 Feb 12
18:17 CET


What a surprise to see that there was a beautiful background under that black paint. This painting is finally getting it's due recognition. The colors she is wearing are of the Sforza-Visconti dynasty, proving she is of noble blood.

6 Feb 12
17:29 CET


Why does the media say this painting is just now discovered, when it has long been one of the better known copies out there? But, glad, this painting is finally getting the recognition it deserves. She is seen wearing the Sforza-Visconti colors of red, black and white, giving us clues to her identity as told to us by historian Maike Vogt-Luerssen. Maybe the world will become one step closer to accepting her true identity, and believeing her and Leonardo's amazing clandestine history. Now, when will the the Italian copy that actually looks like it could be painted by Leonardo get recognition get "discovered"? Will the world accept her true identity or will it continually be suppressed by influential 'DaVinci' experts that are less than willing to admit this mystery was already solved by the research of Mrs. Vogt-Luerssen? A silk merchant's wife she was not!!

6 Feb 12
17:26 CET


Copying Leonardo's paintings by his students was an accepted practice in the 16th century. The original painting by da Vinci included an arch and two columns on either side of the sitter, cut off. Francesco Melzi was Leonardo's heir. When the master died in Amboise, France in 1519, Melizi was given several paintings by his master which he sold to king Francis I, Leonardo's patron. Melzi returned to Milan, a wealthy young man, see my book, "The da Vinci Diaries" Professor James LaMalfa

4 Feb 12
17:20 CET


It is such a joy so see the original setting of the sitter appear in full detail, and we should be grateful that the Prado version was almost forgotten for so long, the painting apparently surviving in an astonishing condition.

4 Feb 12
17:19 CET


Why two copies? The answer might be found in "Leonardo' Val di Chiana Map in the Mona Lisa", in Cartographica, 46:3, 2011. The image on one edge continues on the other edge, forming a new landscape that matches Leonardo's Val di Chiana map. Placing the two copies side-by-side creates a stereoscopic arrangement.

4 Feb 12
17:19 CET


I like the original as the eyes on the Prada copy are crossed and she looks like a typical inbred noble woman

4 Feb 12
17:19 CET


..what's arresting here,it's richness of detail (indicating on taking from nature),bigger and closer to Vasari's description. One question remains:Is there anything under that landscape reminding Rafael's "Mona Lisa" - his drawing from that period (~1503) ?

4 Feb 12
17:19 CET


1) Not a 'delicate tuscan landscape'. There's nothing like that in Tuscany. And what we see here is certainly not 'delicate'. Rather a harsh and rocky landscape. 2) Salai was NOT Leonardo's lover. The fact that there's no evidence of Leonardo's sexual life is not enough to say that Salai was his lover. He was his assistant and pupil, more like an adopted son.

3 Feb 12
14:58 CET


I like the face on the new version better. It seems more realistic. The hands on the Leonardo version are better, though, so neither one is better in all respects. It has been pointed out that the faces on women in many of Leonardo's paintings look alike. This might be because Leonardo forced the people in his paintings to conform to his theories of ideal human symmetry. That is one reason why I think the face on the new version is closer to what the woman actually looked like, but also, the face just looks more real in some indefinable sense. Because the face seems more real, it is more engaging to me, because I can empathize with it more. Perhaps the author of this article is being diplomatic when he says the cracked varnish makes Leonardo's version look middle aged. The puffier face, and particularly the swollen, droopy eyebrows are not the result of varnish.

2 Feb 12
23:53 CET


What an extraordinary discovery, indeed. I must admit, looking at the image of the copy, I can not resist the feeling I find it almost more attractive than the original. The coming show at the Louvre is definitely worth a visit...

2 Feb 12
15:10 CET


Thanks for posting this very revealing story--and I love not only the eyebrows but the catchlights. I've written an essay on Mona Lisa where I have always maintained that the original had those catch lights, plus many more details.

2 Feb 12
15:9 CET


good news i always wanted MONA LISA get the proper credits that she was the woman in the portrat.

2 Feb 12
15:9 CET


This work from the Prado was used in my Mona Lisa Restoration Project, a high resolution, full color fine art recreation of Leonardo's original work.

2 Feb 12
15:8 CET


I can't get past the apparent picture of the face of Ghengis Khan hidden in the mountain, a bit under the peak...

Submit a comment

All comments are moderated. If you would like your comment to be approved, please use your real name, not a pseudonym. We ask for your email address in case we wish to contact you - it will not be made public and we do not use it for any other purpose.


Want to write a longer comment to this article? Email


Share this