The filming of Dr. Strangelove: Shepperton, locations and work organisation

The organisation of the filming of Dr. Strangelove brought together constructed interiors, exterior footage and a cast required to sustain different comic registers.

The British production of an American story

Dr. Strangelove is an American film in its characters and much of its conflict, but its production took place in the United Kingdom. Shepperton Studios occupies a central place in accounts of the filming. This difference between the production location and the represented setting is especially important: the War Room, the airbase and the B-52 cockpit are narrative locations, not a route that allows the filming sites to be identified directly. The work dates belong to the 1963 production and must be distinguished from the 1964 commercial release.

The scale of the project required sets, actors, cameras and footage intended to represent exteriors to be coordinated. The large interiors imposed different lighting and movement requirements from the enclosed cockpit scenes. A shooting schedule organises these demands according to the availability of equipment and cast, not necessarily in the order of the plot. Without dated production sheets, an exact day-by-day schedule cannot be reconstructed. The B-52 set and the aerial exteriors make it possible to examine separately two types of work that converge in the finished film.

Actors, takes and accounts

The cast brings together performances in very different registers. Peter Sellers plays President Muffley, Captain Mandrake and Dr Strangelove; Sterling Hayden gives General Ripper gravitas, while George C. Scott plays Turgidson with an energy that shifts between the military and the grotesque. The filming had to allow these registers to coexist within the same film. Accounts of repeated takes, improvisation or differences of opinion are useful, but they should be attributed to the person who told them. A popular anecdote alone is not enough to prove widespread conflict on set.

Film comedy also depends on what is selected afterwards. A performance may offer variations that only acquire their definitive function in the edit. That does not make every scene an improvisation, nor does it allow us to infer how actors were directed from a single surviving take. Production photographs can document positions and objects, but not the duration or context of decisions. The production dossier deals with industrial arrangements; here, the focus is on the day-to-day work that turns a script and a collection of designs into performed images.

Peter Sellers’s multiple roles pose a particularly interesting organisational problem. His characters belong to different spaces and need not share a scene for the audience to recognise a continuity of tone. Filming can organise his appearances according to sets and availability, while the edit distributes them throughout the crisis. The film uses this separation to let the actor adopt contrasting registers without the audience having to witness a transformation in real time. The logistics and the final form are related, but they are not identical. Without production reports, it is impossible to state the exact order in which his roles were filmed. What can be verified is the difference in performance and its subsequent integration.

Archival images of the filming can also lead to errors of location. A photograph taken at Shepperton may depict a fictional American place, while an exterior incorporated into the edit may come from footage obtained by another unit. Identifying a real site requires documenting the source and date, rather than inferring it from the landscape or an actor’s uniform. This distinction matters when compiling production chronologies or maps. Filming is an industrial activity spread across departments and time; the finished film is a constructed continuity. Describing these two scales separately makes it possible to understand how the work was organised without treating the images as evidence of a non-existent geography. Studio photographs also reveal the collective nature of the work: a seemingly self-contained set depends on cameras, lights and staff positioned out of shot. This distinction between the filmed space and the production space should be retained when reconstructing the history of the filming.

Sources for reconstructing the logistics

Production documentation can include shooting reports, contracts, photographs, interviews, correspondence and later memoirs. Each source has a different scope. A credit confirms a professional role; an interview may describe an experience; a photograph records a moment. No single item is enough to reconstruct every unit, location and decision. In Dr. Strangelove, it is particularly easy to confuse the fictional space with the filming location, because the film depicts military installations with great visual force. Accuracy requires these two geographies to be kept separate.

Research into the filming must also resist turning Kubrick’s reputedly demanding nature into an automatic explanation for every disagreement. A complex production may involve schedule adjustments, design changes or additional takes for many different reasons. Describing a specific incident requires names, dates and accounts that can be corroborated. What the film allows us to state with confidence is the result of coordination between departments: convincing spaces, performances in varying tones and a continuity that makes the satire possible. The story of how this result was achieved should be based on documents, not the director’s later reputation.

View all articles in Dr. Strangelove: characters, mise-en-scène and nuclear satire