Drawings by Jan van Eyck

Jan van Eyck’s drawings and working methods

The surviving drawings associated with Jan van Eyck range from a closely observed portrait study to sheets made decades after his death. They preserve different kinds of evidence: a sitter's appearance, a composition used for transfer, or a later artist's response to an admired model. The Dresden Portrait of an Older Man is the secure starting point. The Morgan describes it as the only drawing universally accepted as Jan's own work. That judgment concerns the surviving sheet, rather than every design connected with his name.^[1]

This guide follows the objects and their relationships. The Dresden portrait study and Rotterdam Crucifixion have separate pages for their detailed histories. Here they are compared with workshop and circle drawings, copies after lost paintings, and preparation beneath surviving paint. An attribution to Jan, a connection with his workshop and a copy after his invention are different claims. Their distinctions make this small group more informative than a longer list that treats every Eyckian image as his autograph work.^[1][2]

Drawing with a metal point

A metalpoint drawing depends on its prepared surface. A stylus deposits tiny particles as it moves across an abrasive ground; the image is made through contact between tool and coating. An Van Camp's survey for CODART describes fifteenth-century paper prepared with bone ash and animal glue. Pale grounds were particularly common in the Netherlands. The deposited line may change colour as the metal reacts with its surroundings, so the present brown or grey appearance does not necessarily reproduce what the draughtsman first saw.^[2]

The technique permits very fine contours and close hatching. It also demands care: unwanted marks are difficult to remove, and pressure can damage the ground. Those properties suit precise observation, but they also suit the preservation of figures and designs for reuse. Van Camp stresses that many surviving fifteenth-century metalpoint drawings are workshop copies. The material therefore cannot establish an autograph attribution by itself. A finely finished metalpoint sheet might record a live sitting, copy an existing picture or provide a model for another maker.^[2]

Paper drawings also need to be distinguished from lines on other supports. The Met's Saint Paul is drawn on vellum, while Saint Barbara is on panel. A prepared paper sheet could be handled separately from a painting; an underdrawing on a panel belongs to the making of that painted object. The word “drawing” can describe all their linear effects in ordinary conversation, but a collection record must identify which physical thing survives.^[3][4]

The Dresden portrait study

Dresden's Kupferstich-Kabinett holds the Portrait of an Older Man, inventory C 775. Silverpoint and goldpoint model the face on white prepared paper. The closely packed strokes describe the planes of the head and the changes of light across the skin. Written notes preserve information about colouring that those monochrome marks could not carry. The Morgan links the sheet directly to the painted likeness in Vienna and questions its traditional identification as Cardinal Niccolò Albergati, proposing an unidentified scholar instead.^[1]

That relationship offers an unusually direct view of portrait preparation. The face could be recorded during an encounter and its colours retained in words for later use. Contour, shading and writing do different jobs on the same support. The notes were working information, rather than a caption identifying the sitter for posterity. The resulting combination helps explain why a vivid surviving likeness can be more securely connected with a painting than with a named historical person.^[1]

The approximate date has changed between publications. CODART's 2015 survey gives about 1432, while the Dresden–Morgan exhibition uses about 1435–40. The National Gallery's material for its 2026 portrait exhibition dates the sheet about 1438 and describes the related Vienna painting as an elderly man, possibly a scholar. These are curatorial estimates, not three dated sittings. The recent wording strengthens the case for keeping the familiar Albergati title qualified and for distinguishing the drawing's secure authorship from its less secure personal identification.^[2][5]

Technical research has also complicated the idea of one uninterrupted session with a single tool. A scholarly review of the Dresden exhibition catalogue reports two silverpoint types and a goldpoint, used for different passages of the face, surrounding shading and inscriptions. The review is evidence for the published investigation, rather than a substitute for inspecting its complete scientific data. At guide level, the result establishes a varied use of metalpoint; it does not give a clock for the sitting or prove the exact order of every mark.^[6]

Rotterdam's Crucifixion and the attribution debate

The Rotterdam sheet measures 25.4 by 18.7 centimetres and is catalogued as goldpoint, silverpoint, pen and black ink on prepared paper. Museum Boijmans Van Beuningen records its acquisition in 2014, following its appearance in The Road to Van Eyck. Its history of modern attention reaches back further: Till-Holger Borchert reports that the owner brought it to Bruges for examination in 2002. “Rediscovered” can describe the moment of public recognition without implying that nobody had previously inspected the object.^[7][8]

Boijmans currently displays Jan's name in the maker field, also lists an anonymous maker in the specifications, and dates the sheet about 1440–50. The later part of that range falls after his death in 1441. These entries do not resolve the debate; they preserve an awkward but useful indication of it. The picture is closely related to the Met's painted Crucifixion, yet its broader composition and altered groups prevent a simple identification as a reduced version of the surviving narrow panel.^[7]

Arie Wallert's technical investigation identifies a bone-white ground and different metal alloys. It also reports blue indigo on the reverse and impressed contours consistent with tracing. These observations concern making and use: the sheet could serve as a model whose outlines were transferred. Wallert argues that its design preceded the New York painting and belongs very close to Jan's working practice. He does not identify a surviving product made by directly tracing this sheet.^[9]

Borchert's Dutch-language response accepts that the discovery of metalpoint changed an earlier, very late assessment. He nevertheless argues for a later fifteenth-century copy after an older Eyckian composition. Goldpoint and silverpoint were period practices, rather than signatures belonging to one artist. In this reading, shared motifs connect the sheet with a lost design or a network of models, while the surviving drawing's authorship remains separate. The two interpretations address the same object but give its marks different places in the sequence of invention, copying and reuse.^[8]

The difference matters because a working model need not be the first drawing of its composition. A sheet can be copied and then used to make further copies. Physical traces of transfer establish activity around that particular support without automatically identifying the inventor or the recipient. The Rotterdam object page examines the technical arguments in full. Here the sheet remains a serious source for Eyckian workshop practice with a disputed personal attribution, rather than a second universally accepted autograph drawing.^[8][9]

Watch: Jan van Eyck in context

Choose a Smarthistory close reading of the Arnolfini Portrait or the closed and opened Ghent Altarpiece.

Jan van Eyck, The Arnolfini Portrait

Smarthistory

Look closely at the room, mirror and gestures in the London double portrait.

Watch on YouTube

▶01 · PortraitThe Arnolfini PortraitSmarthistory▶02 · ExteriorThe Ghent Altarpiece, closedSmarthistory▶03 · InteriorThe Ghent Altarpiece, openSmarthistory

Looking beneath the paintings

The Met brought the Rotterdam sheet together with its painted Crucifixion in 2016. Infrared reflectography enabled comparison with underdrawing beneath the painted pair, while X-radiography revealed fragmentary wording hidden on the frames. Those methods investigate different layers and materials. Similarity between a paper drawing and a painted underdrawing can clarify design relationships, but it is not a dated record establishing that one object was made immediately before the other.^[10]

The date of the painted comparator also needs to be identified by publication. The Met's 2016 exhibition used about 1440–41 and a Jan-and-workshop attribution; its current collection record gives Jan van Eyck and about 1436–38. The older research remains useful, but its label cannot simply be treated as current. A proposed sequence involving the Rotterdam sheet should specify which chronology it is testing.^[10][11]

Underdrawing varies within Jan's paintings. The National Gallery describes relatively little in the possible self-portrait of 1433, with evidence of adjustments during painting. A meticulously described surface does not therefore imply a complete separate paper plan for every visible detail. Looking at concealed preparation helps recover decisions made on a panel, while the absence of a corresponding sheet leaves open how much planning took place elsewhere. This evidence expands the account of drawing practice without enlarging the surviving paper corpus.^[12]

The Louvre’s Fishing Party: a changed object

A further disputed candidate is the Louvre’s Une partie de pêche, usually called the Fishing Party, inventory INV 20674. The collection record gives about 1420 while retaining “copy after” Jan in its main attribution. Its discussion records proposals by Claudine Chavannes-Mazel and Pierre Colman to recognise Jan’s own hand, alongside other assignments to an early Netherlandish maker. It also cites a 2024 catalogue discussion. These different judgments should be stated together; the date field cannot quietly settle the maker field.^[13]

The surviving image combines groups of richly dressed men and women with water and a distant setting. Its medium includes pen, wash, watercolour, paint and gold and white highlights. That mixture distinguishes it from the monochrome portrait study. It also makes present appearance a particularly unreliable guide to the initial balance between line and colour. A work can now look faint or incomplete because its materials have altered, rather than because its maker stopped at a preliminary stage.^[13][14]

Chavannes-Mazel and her collaborators’ technical study identifies two fragments from a once-coloured composition, joined with a central portion missing. Elemental imaging helped recover evidence for faded colour. The authors argue that later additions turned hunting sticks into fishing rods and repainted parts of the setting. The familiar title thus describes a modified image. Their reconstruction places the groups beside a larger body of water, rather than facing one another across the narrow channel suggested by the present join.^[14]

The case adds a useful qualification to any claim that Jan’s drawn corpus is settled. Strong arguments have been made for expanding it, but the proposal is not equivalent to the broad acceptance of the Dresden head. Here, questions about the original subject, missing paper, later repainting and authorship remain linked. A reliable guide must allow the sheet into the discussion without converting its most ambitious attribution into an agreed second autograph work.^[13][14]

Later sheets after lost compositions

The Louvre's Saint Christopher Carrying the Christ Child, INV 20664 recto, is dated about 1480 and catalogued as a copy after Jan. The museum points to a watermark not attested before the 1470s. That material evidence supports a late date for the surviving paper even if the design has an older ancestry. The attribution “after” connects the sheet with an invention; it does not turn its later brush and wash into marks made by Jan.^[15]

The Louvre also records disagreement about how much of the image derives from a lost original. One interpretation assigns only the saint to the Eyckian model and associates the landscape with another artistic context. The relationship is thus more specific than a general resemblance but less secure than a complete reconstruction. The Saint Christopher page compares the drawing with the later Philadelphia painting. Each witness must retain its own date, medium and history.^[15]

At the Albertina, Christ Carrying the Cross, inventory 3025, is assigned to Dieric Bouts's workshop about 1460–70. Brown pen work lies over a grey preliminary drawing. The curatorial account connects its crowded procession with a proposed lost Van Eyck scheme of about 1420–25, transmitted through copies. Some passages, including the group with Veronica and the mourners, were not completed in ink. The museum interprets the contrast between carefully executed and omitted areas as a sign of copying.^[16]

The Vienna sheet is therefore evidence of a later workshop's selective use of a composition. The proposed early date belongs to the lost model, not to the paper now in the Albertina. Comparison with related painted versions reveals changes in the direction of the route, the landscape and the chosen narrative moment. Those differences make the drawings useful for studying transmission, while limiting their ability to restore a vanished original exactly.^[16]

A silverpoint witness to the lost Maelbeke Virgin offers a further example. The Frick's exhibition attributes this sheet to Petrus Christus, after Jan and his workshop, about 1445. It preserves the Virgin and Child and the architecture while omitting much of the landscape and donor imagery. Its omissions are part of the evidence: a later artist could select what was useful rather than reproduce every feature. The sheet documents an encounter with an Eyckian composition, not the survival of Jan's own preparatory drawing.^[17]

A circle drawing on vellum

The Met's Saint Paul, about 1430, is catalogued to the circle of Jan van Eyck. It was formerly attributed to Petrus Christus. Pen and brush in brown ink, with purple and gold details, describe the figure on vellum measuring 14.6 by 7.9 centimetres. The Met suggests that the support's translucency might have permitted some straight contours to be traced from a model, with shading added afterwards. It retains that process as a possibility.^[3]

The museum relates the figure and its prototype to painting, possibly grisaille, rather than sculpture. “Circle” nevertheless identifies an artistic relationship without establishing that Jan drew, designed or supervised this particular sheet. The distinction is useful beside the Maelbeke copy: one label proposes a milieu, while the other identifies a dependence on a particular lost composition. Neither needs promotion to an autograph work to contribute to the study of Eyckian drawing.^[3][17]

Panel and manuscript boundaries

Saint Barbara is dated 1437 on its original frame and catalogued by KMSKA as oil on panel. Its detailed linear work and restricted colour have prompted interpretations ranging from an abandoned painting to a deliberately monochrome object. Even the imitation marbling on the frame is incomplete in one corner. Its drawn appearance is revealing, but its support and disputed finish prevent it being counted as another loose sheet beside Dresden and Rotterdam.^[4]

The Birth of Saint John the Baptist belongs to the Turin–Milan Hours. Its proposed attribution to Jan as the illuminator called Hand G remains disputed. Carol Herselle Krinsky argues for a later campaign after his death, challenging the manuscript's assumed connection with John of Bavaria. A manuscript miniature can illuminate the reach of Eyckian ideas, but should enter a drawings guide with its medium and contested attribution visible.^[18]

Reading the small surviving group

Each object answers a different question. Dresden connects observation and written colour notes with a painted portrait. Rotterdam tests competing accounts of invention and transfer. The Louvre, Albertina and Maelbeke witnesses show designs being preserved and changed by later makers. The Met's Saint Paul provides evidence for an artistic circle without identifying Jan's hand. The useful comparison follows those relationships rather than treating physical beauty or a familiar subject as sufficient proof of authorship.

The collection record is the starting point: identify the support, the date of the surviving object and the precise attribution. Then separate that evidence from the date or maker proposed for a lost model. Scientific examination can identify materials and reveal use; historical and stylistic comparison must still explain what those findings mean. Keeping those tasks connected allows a very small secure corpus to support a substantial account of drawing, without turning the wider Eyckian tradition into an invented collection of autograph sheets.

Related reading

Jan van Eyck’s Dresden portrait study, traditionally called Cardinal Albergati

The Crucifixion Drawing Associated with Jan van Eyck

Saint Christopher: tracing a lost Eyckian design

References

1. Morgan Library & Museum, “Jan van Eyck: Portrait of an Older Man,” Dresden–Morgan exhibition: attribution, materials, observation and questioned sitter. View Source
2. An Van Camp, “Metalpoint Drawings from the Low Countries,” CODART, June 2015: materials, workshop functions and portrait study. View Source
3. Metropolitan Museum of Art, “Saint Paul,” 1975.1.841: circle attribution, vellum and possible tracing. View Source
4. KMSKA, “Saint Barbara,” 410: panel support, 1437 frame and disputed finish. View Source
5. National Gallery, London, “Van Eyck: The Portraits,” 2026 exhibition announcement: Dresden study dated about 1438 and related Vienna portrait. View Source
6. Historians of Netherlandish Art, review of Das Geheimnis des Jan van Eyck, November 2007: report of technical investigation of the Dresden sheet. View Source
7. Museum Boijmans Van Beuningen, “Crucifixion of Christ,” MB 2014/T 1 (PK): materials, maker fields, dimensions, acquisition and exhibition history. View Source
8. Till-Holger Borchert, “Een Eyckiaanse tekening op de expo De weg naar van Eyck,” Flemish Art Collection: Dutch-language attribution and dating argument. View Source
9. Arie Wallert, “Function and Meaning of a Metal Point Drawing by Jan van Eyck,” ArtMatters 5 (2013), full article: examination and proposed working function. View Source
10. Metropolitan Museum of Art, “A New Look at a Van Eyck Masterpiece,” 2016 exhibition research. View Source
11. Metropolitan Museum of Art, “The Crucifixion; The Last Judgment,” 33.92ab: current attribution and dating. View Source
12. National Gallery, London, “Portrait of a Man (Self Portrait?),” NG222: limited underdrawing and changes during painting. View Source
13. Musée du Louvre, “Une partie de pêche,” INV 20674: composite drawing, mixed media and competing attribution proposals. View Source
14. Claudine A. Chavannes-Mazel with Matthias Alfeld, Koen Janssens, Geert Van der Snickt and collaborators, “The Fishing Party by Jan van Eyck (?). A Technical Analysis,” Van Eyck Studies (2017), author-posted article: fragments, pigment loss and later additions. View Source
15. Musée du Louvre, “Saint Christophe passant l'Enfant Jésus,” INV 20664: copy after Jan, watermark and interpretation of the lost model. View Source
16. Albertina, “Kreuztragung Christi,” 3025: Bouts workshop, about 1460–70, materials and proposed lost Eyckian design. View Source
17. Frick Collection, “The Virgin and Child with a Donor,” Charterhouse exhibition: selective Maelbeke copy attributed to Petrus Christus. View Source
18. Carol Herselle Krinsky, “The Turin-Milan Hours: Revised Dating and Attribution,” Journal of Historians of Netherlandish Art 6.2 (2014): disputed Hand G identity and manuscript chronology. View Source