Art historian and author John R. Eyre, in his Monograph on Leonardo da Vinci’s Mona Lisa (1915), and also in his The Two Mona Lisas (1926), was a dedicated believer that “Leonardo … worked up two versions of the same picture at the same time; in this case he began two portraits of Lisa del Giocondo around 1501. One was consigned to Francesco del Giocondo [her husband], while the other – the painting now in the Louvre – remained unfinished until Giuliano de Medici asked that it be completed.” Eyre is adamant that the ‘Isleworth Mona Lisa’ is the one painted for Giocondo: both of his books focus entirely on this.

Léon Roger-Milès, in his Leonard de Vinci et les Jocondes (1923), also contends that Leonardo painted two Mona Lisas (and possibly more): the first for Francesco del Giocondo, and the second for Giuliano de Medici.

In an article entitled La Joconde et Les Jocondes, the author Guy Isnard writes: “ … We can therefore ask ourselves if there were not at least two ‘Giocondas’ painted by Leonardo; the first at the request of Francesco del Giocondo, that was seen by Raphael in 1504 at Leonardo’s studio in Florence, incomplete and of which he made a sketch; the second, painted in 1513 at the demand of Giuliano de Medici, when Monna Lisa was his mistress*. The canvas was presented on October 10, 1517 by the painter, to Cardinal Luigi of Aragon and to his secretary, Antonio de Beatis, at the Château in Cloux.”

*Editor’s note: Though Lisa may have been acquainted with some persons of the Medici family, there is no evidence to prove that Lisa del Giocondo was anything other than a loyal and faithful wife and mother, and that she lived her whole adult life in Florence.

In his 1960 book, old master specialist Henry F. Pulitzer strongly agrees with Eyre that Leonardo painted two versions, and that the earlier one is the ‘Isleworth Mona Lisa’. With regard to the painting in the Louvre however, Pulitzer sides with the earlier theory of Adolfo Venturi, that she is Costanza d’Avalos, Duchess of Francavilla.

Frank Zöllner (Leonardo’s portrait of Mona Lisa del Giocondo, 1993) after researching the lack of flanking columns, or pillars, in the Louvre ‘Mona Lisa’ comments: “This new evidence seems to support the suggestion that Leonardo painted two versions of Mona Lisa, yet the possibility of a cartoon version with full columns is more likely.” Since there exists many copies at much later dates it seems hardly likely that these would be based on a cartoon when the original painting was clearly finished. So it is far more credible that both Raphael and later copyists based their work on an earlier original version.

The Louvre Museum itself in a 2017 publication authored by Vincent Delieuvin admits that Raphael and other copyists must have based their work on a version on Mona Lisa by Leonardo other than the one today in the Louvre.