Looking at a Rubens painting
Peter Paul Rubens painted altarpieces, portraits, mythological scenes, landscapes and designs for vast court projects. An encounter with any one of them begins with a particular object: its support, size, present collection, original purpose and attribution. “A Rubens painting” can mean a work painted substantially by the artist, an oil sketch for a larger commission, a picture completed with collaborators, or a studio painting after his design. Those distinctions make looking more rewarding because they explain what the surviving surface can actually tell us.
The image above is the Museo del Prado's Saint George and the Dragon, catalogued as an oil painting on canvas from 1606–08. George's armoured figure rises on a white horse over the dragon, while the princess and a lamb stand nearby. It gives a vivid entrance to Rubens's interest in bodies in motion and narrative suspense during his Italian years. The object page treats its identity and setting in detail. Here it introduces a broader guide: how to move among Rubens's painted works without losing sight of their different materials and histories.
At a Glance
- Local painted-object routes101
- Other drawing routes27
- Image aboveSaint George and the Dragon, Prado P001644 [3]
- AttributionCheck each object's current collection record [1][2]
Browse all 101 Rubens paintings
What counts as a painting in this guide?
A finished picture can be made on panel, canvas, slate or another support. A small oil sketch is painted too, although its task might be to work out a composition, propose colours or show a patron the plan for a larger object. Calling both “paintings” describes their medium, not the same stage of production. The surface of an altarpiece designed for a church needs to be understood at its architectural scale; the surface of a cabinet picture invites close viewing. The guide includes both when the route concerns a painted object, and labels the specific kind of object on its own page.
This matters because the works did not all travel through the same process. The Metropolitan Museum describes Rubens developing compositions in drawings, oil sketches and detailed figure studies, then using assistants for some large commissions. Its account identifies a common studio practice, but no single sequence should be imposed on every work. A finished canvas may preserve revisions, or a sketch may be a fully expressive object in its own right. The oil sketches and technique guide follows that process more closely; the drawings guide covers works on paper that belong outside a painting-only directory.
The large Corpus Rubenianum Ludwig Burchard is often invoked when people ask how many works Rubens made. Rubenshuis says it charts more than 2,500 compositions and over 16,000 related works, explicitly including drawings, sketches, prints, tapestries and copies by pupils or successors. These are documentation figures across a very broad scholarly catalogue. They do not mean that Rubens personally painted 16,000 pictures, and they should not be used as a total for the painted objects listed here. The directory below is this site's set of checked local routes, not a complete catalogue of everything made in his lifetime.
Altarpieces and religious histories
Church commissions required pictures able to communicate across large spaces and within established devotional programmes. Rubens's Elevation of the Cross and Descent from the Cross in Antwerp Cathedral are triptychs, not simply two rectangular images. Their central scenes join painted wings that change how the story unfolds. The Descent, made in 1611–14, was made for a specific place and patronage context. A reproduction of its centre cannot show the complete relation between panels or the physical experience of standing before it.
Religious stories also appear in very different formats. Daniel in the Lions' Den in Washington places a solitary figure among animals in a large horizontal painting. The National Gallery of Art dates it about 1614–16 and identifies nine lions. Its suspense comes from the animals' varied positions and Daniel's raised gaze, not a crowd of combatants. By comparison, the Assumption of the Virgin above the high altar at Antwerp Cathedral answers a different spatial and liturgical need. Both are religious histories, but that shared category does not erase their distinct settings, supports or audiences.
Painted religious pictures could also change as a commission developed. Rubens made an initial altarpiece for Santa Maria in Vallicella in Rome; that work is now at the Musée de Grenoble. A later solution remains associated with the church. The two should not be treated as one object that simply moved from Rome to France. The early Saint Gregory with Saints and the Roman Madonna della Vallicella pages identify their respective materials and locations. Their relationship shows why version and site are part of a painting's identity.
Myth, literature and the moving body
Classical stories gave Rubens scope to arrange bodies in action, but their painted forms varied from intimate panels to monumental canvases. In Prometheus Bound at the Philadelphia Museum of Art, the figure twists under the attack of an eagle. The eagle is associated with Frans Snyders, an artist known for animal painting. That named contribution is evidence for this picture, not a general rule that Snyders painted every creature in a Rubens scene. The whole painting depends on the forceful meeting of human body, bird and rock, whatever division of labour produced its passages.
Rubens's Hippopotamus and Crocodile Hunt in Munich puts animals and riders into a different kind of violent encounter. It was part of a group for Elector Maximilian I of Bavaria, according to the Alte Pinakothek. Hunting pictures could present spectacle for a princely audience, while myths could turn a familiar ancient narrative into physical drama. The viewer should still ask what the collection record documents about a given work's patron or series before treating all energetic animal paintings as one commission.
Late mythological painting could be quieter. The Prado's Three Graces is oil on oak, a material point easily lost when the image is reproduced on a screen. Three linked figures create a rhythm of joined hands, turned bodies and light. Its late date invites comparison with portraits of Rubens's family, but visual resemblance alone cannot establish the identity of each figure. The page treats the painting as a specific panel with its own provenance and subject, rather than using the artist's marriage as a shortcut to interpretation.
Portraits, family images and court display
Rubens painted rulers, patrons, associates and members of his family. A portrait may record a sitter's features, but it can also state office, wealth, kinship or status. The equestrian portrait of the Duke of Lerma belongs to Rubens's early Spanish court visit. Its horse, armour and elevated viewpoint do political work as well as descriptive work. Later portraits made for courts need to be read against their own commissions and surviving versions, rather than gathered into a timeless gallery of faces.
The portrait of Rubens and Isabella Brant in the Honeysuckle Bower offers a different form of self-presentation. The Alte Pinakothek dates it about 1609–10. The pair sit close together in carefully chosen dress and setting; the image is personal and public at once. It is useful to compare it with Rubens's later self-portrait and with portraits of relatives, while keeping the collection's date and physical object in view. A family relation does not make every portrait an informal private sketch.
Attribution is especially important for portraits circulating under a famous name. A museum may catalogue a painting as “studio of Rubens” or identify later retouching, whereas an inherited page title simply says “by Rubens.” The Norton Simon Anne of Austria illustrates the need to follow the current object record and its wording rather than a headline alone. A drawing for a portrait and the finished oil may also occupy separate pages. The painted-object directory below excludes sheets that are catalogued as drawings, even when they represent the same sitter.
Landscape and the late career
Rubens's landscapes are more than backdrops for figures. The National Gallery's A View of Het Steen in the Early Morning connects the estate he bought in 1635 with a broad scene of fields, buildings, light and people at work. The ownership gives historical context; it does not turn the painting into a precise map of every path. The Rainbow Landscape likewise asks a viewer to follow changing weather, animals, fields and distance across a painted space. These late works reward attention to how light organises the whole picture, not only to the name of the place in a title.
Other landscapes incorporate narratives. Landscape with Saint George and the Dragon in the Royal Collection differs from the Prado Saint George at the top of this guide. One integrates the episode into a broad countryside; the other concentrates on the mounted confrontation. Similar characters do not make two paintings versions of the same physical object. Their separate collections, dates, media and compositions matter. The comparison helps reveal how a subject changes when landscape becomes a major part of the picture's purpose.
Paintings made for courts and political rooms
The Marie de' Medici cycle is a particularly large case of paintings designed as an ensemble. For the Luxembourg Palace in Paris, Rubens developed images of the French queen's life in which historical people meet gods and personifications. The Louvre describes 21 narrative canvases and three portraits in the completed queen's gallery. The scenes depict events from different years; the paintings themselves belong to the 1622–25 project. A single canvas removed from its sequence can be read closely, but the political argument becomes clearer across the gallery.
Rubens's English court work gives another example. Minerva Protects Pax from Mars, often called Peace and War, belongs to his visit in 1629–30. Its figures give form to an argument about peace, abundance and the dangers of war. The Banqueting House ceiling was a separate commission for Charles I. Its planning and installation should not be collapsed into the moment Rubens was in England. These works were made for rulers and rooms with political meaning; the diplomacy and courts guide follows the negotiations around them.
Some great projects are known through multiple types of painted object. An oil sketch may propose a tapestry composition; a larger painted model may guide weavers; a finished tapestry uses thread rather than paint. The Triumph of the Eucharist designs show why a painted study belongs in a painting guide while the woven result has another material identity. A title shared between stages does not justify moving the size, date or technique of one stage onto another.
Versions, copies and attribution labels
Two works in this cluster are called The Massacre of the Innocents. The Art Gallery of Ontario catalogues the early Toronto panel about 1610, oil on panel, object 2014/1581. The Alte Pinakothek catalogues a later Munich work about 1638, on oak, inventory 572. They are separate paintings made decades apart. Their surfaces, dimensions, composition and histories must not be merged merely because an English title matches. The relevant page names the actual collection and version before interpreting the scene.
Attribution labels carry similar weight. “By Rubens” assigns the work to the artist under the collection's current judgement. “Rubens and [another artist]” names a collaboration; it does not automatically divide the surface equally. “Studio of Rubens” places production within his workshop, while “after Rubens” usually signals a later work following his design. “Attributed to Rubens” states a qualified judgement. These are useful distinctions for readers, not small-print obstacles. Different museums and scholars can revise them as evidence changes; each route should report its own record and explain genuine disagreement.
Digital images can conceal these differences. A photograph of a panel does not show its thickness or later additions. A cropped image can omit a triptych's wings. A familiar online image may belong to a copy, not the museum's original. On each artwork page, the caption states what the local image depicts and the source record that supports it. The inherited file name and route remain convenient addresses, but they are not proof of authorship, medium or identity. Browse the painting routes below by the object named in the row, then open a page for its evidence.
From first idea to painted surface
A finished Rubens painting often stands at the end of a chain of other objects, but the chain is not identical for every commission. The Metropolitan Museum's account of his works on paper[15] describes quick pen-and-ink ideas, painted models that could be presented to patrons, and chalk studies of particular figures. For some canvases, Rubens then set out the composition in dark brown oil paint before he or assistants developed the painted image. This account gives a way to ask questions of a work: was its design proposed to a patron, did an assistant transfer a studied figure, and does the finished surface preserve a change of mind? The museum explicitly cautions that he did not repeat one procedure mechanically.
The surviving evidence can make that process visible. In the Descent from the Cross, the Courtauld's small oil study belongs to the preparation of the enormous Antwerp triptych. The Met's ink drawing for its Presentation in the Temple wing shows Simeon and Anna in positions different from those of the final panel. These are changes between distinct objects, rather than changes inferred merely from a photograph of the finished altarpiece. In the Prado's Three Graces, conservators identified extensions to the oak panel at the left, top and right while Rubens was working. There the support itself preserves a revised design. One case concerns an altar project developed through studies; the other concerns a panel enlarged during painting. Neither licenses a universal recipe for Rubens's production.
Material study can also prevent an apparently simple comparison from becoming misleading. The Prado's Three Graces is a very large oak panel. The Garden of Love is a different object whose composition brings together a social gathering, sculpture and amorini. Digital reproductions may make both appear like similarly sized rectangles, while their real supports, scale and histories differ. The same applies to the Antwerp triptychs: their wings and changing open and closed states are part of the objects even when a search result displays only a central panel. A useful first step is to read the holding collection's medium and dimensions before using an image to compare gesture, light or subject.
The studio and what an attribution can say
Rubens's studio was capable of more than copying completed pictures. The Met's drawings exhibition[2] describes how the artist kept drawings as working material for pupils and collaborators. A figure study could guide someone painting part of a larger commission; Rubens might subsequently retouch the work. The museum also records that in 1611 he said he had turned away more than a hundred prospective pupils. That statement speaks to demand for training, not to the number of artists employed on a particular canvas. It is better evidence for the studio's reputation than for assigning any individual hand in a painting.
Named specialists show a more specific kind of collaboration. Frans Snyders is associated with the eagle in Philadelphia's Prometheus Bound. Jan Brueghel the Elder worked with Rubens on pictures such as Pan and Syrinx, where figures and detailed natural setting come together. Those associations have to be established for each object. A broad claim that Rubens delegated every animal or all landscape passages would conceal both his own range and differences among commissions. Conversely, a joint credit does not diminish the work's significance. It records how distinct skills were brought together on one surface.
The Getty's study of Rubens and Brueghel[16] compares their handling in works studied technically. It describes Brueghel's more opaque, detailed application and Rubens's frequent use of translucent glazes and scumbles to build modeled form. It also notes Rubens's use of warm brown paint to lay in and adjust contours, and instances where a prepared ground remains visible in flesh shadows. These are observations from examined works, not an instruction to see a specific pigment or paint layer in every online photograph. They help explain why an attribution is an evidence question: comparison can involve materials, sequence, revisions and documented collaboration as well as resemblance of subject.
Beyond the easel painting
Painted designs may be one stage of an object made in another medium. The Getty's account of the Triumph of the Eucharist[17] describes Rubens planning 20 tapestries for the church of the Descalzas Reales in Madrid from Antwerp, using descriptions of a building he had not seen. Oil sketches carry compositional and colour ideas, while the woven hangings have their own scale, materials and installation. The Getty notes that the arrangement apparently required two levels of tapestries in the church. An isolated painted sketch cannot show that whole spatial programme.
The Met's account of the Constantine tapestries[18] distinguishes small oil modelli by Rubens from larger cartoons produced by pupils and from finished woven panels. This is an especially clear example of why authorship and medium must be specified at each stage. One may truthfully say that Rubens designed a tapestry cycle while still distinguishing his painted design, a studio cartoon, and a textile made by weavers. The paintings guide includes the painted stage where a local route exists; it does not relabel the textile as an oil painting.
How to use the painting directory
The directory lists 101 local painted-object routes whose current checked classification is “Painting.” It includes finished paintings and painted studies, and it can include qualified studio or copy attributions. It excludes 27 drawing routes; those belong in the drawings guide, regardless of whether a drawing shares a subject with a painting. The deliberately retired unidentified image once attached to Last Supper Sketch is not evidence for an oil painting and does not appear in this painted-object list.
Start with a subject that interests you, then compare collection, date, medium and attribution on the individual page. If two titles look alike, check object numbers before assuming they represent the same work. If an artist name is qualified, keep the qualification while reading the scene. The point of the directory is to make Rubens's varied painted production navigable without turning every surviving related object into a personally painted masterpiece.
Browse Rubens painted objects
Frequently Asked Questions
No single number in the scholarly Corpus is a count of paintings entirely painted by Rubens. Rubenshuis's figures of more than 2,500 compositions and over 16,000 related works span drawings, sketches, prints, tapestries, workshop production and copies. This page lists 101 checked local painted-object routes, not a complete lifetime total.
Yes: it is an object made with paint, often on a small support. Its role may be preparatory, such as testing a design or presenting one to a patron. A sketch and a finished canvas of the same subject are separate objects and need separate records.
No. He worked with assistants and named specialists, and later artists copied his designs. Museum labels such as “studio of,” “attributed to” or “after” identify different relationships. The Prometheus Bound collaboration with Frans Snyders is one documented example, not a rule for every animal painting.
Rubens treated the subject at different times. The Toronto panel, about 1610, and the Munich oak painting, about 1638, are distinct objects in distinct collections. Compare their object records and images before transferring a fact from one to the other.
Try the Antwerp Descent from the Cross for church painting, Washington's Daniel for a biblical history, the Prado Three Graces for late myth, and London's Het Steen for landscape. Each page identifies its own object and evidence.
Discussion
Which painted work made you look more closely at its material or purpose?
Share an object record or reliable source if it clarifies a date or attribution.
Reader Insights
How does an oil sketch differ from a finished commission?
Which collaboration changes how you read a familiar painting?
Which separate versions of a subject would you compare?
Join the Conversation
Add a specific observation or source below.
References
- Rubenshuis, Corpus Rubenianum Ludwig Burchard. View source
- Metropolitan Museum of Art, Peter Paul Rubens (1577–1640): The Drawings. View source
- Museo del Prado, Saint George and the Dragon. View source
- Cathedral of Our Lady, Antwerp, The Descent from the Cross. View source
- National Gallery of Art, Washington, Daniel in the Lions' Den. View source
- Musée de Grenoble, Rubens's first Vallicella altarpiece. View source
- Philadelphia Museum of Art, Prometheus Bound teaching resource. View source
- Alte Pinakothek, The Hippopotamus and Crocodile Hunt. View source
- Museo del Prado, The Three Graces. View source
- National Gallery, London, A View of Het Steen in the Early Morning. View source
- Louvre, the Galerie Médicis. View source
- Art Gallery of Ontario, early Massacre of the Innocents. View source
- Alte Pinakothek, late Massacre of the Innocents. View source
- Norton Simon Museum, Anne of Austria. View source
- Metropolitan Museum of Art, Rubens and Van Dyck works on paper, including Rubens's painting sequence. View source
- Getty Research Institute, Rubens and Brueghel: A Working Friendship , technical study. View source
- Getty Museum, The Triumph of the Eucharist tapestry designs. View source
- Metropolitan Museum of Art, Constantine tapestry designs, cartoons and weavings. View source






































 Peter Paul Rubens.jpg)































































