Sydney Harbour Bridge

What it actually is

It is a single steel through arch, and nearly every loose description gets the type wrong. Period engineering sources call it a two-hinged spandrel-braced arch with a clear span of 1,650 feet; the American Society of Civil Engineers, which lists it as a landmark, calls it a steel through arch.

Three points follow, each the opposite of what is usually written. There is one arch, not two: what look like a pair are two parallel braced arch trusses, cross-braced to each other, and each of those is itself a truss with top and bottom chords and diagonal webs. The deck does not run beneath the arch — it runs through it, hung from it on vertical hangers at roughly mid-height, which is what the word "through" in through arch means and what distinguishes it from the suspension bridges it gets compared to. And it is not a tied or anchored arch: the thrust goes into masonry and concrete skewback abutments founded in rock, with no tie member taking the horizontal load.

The hinges are real and the old description of them is sound. It is a two-hinged arch, pinned at the base of each truss where it meets the pylons — four pinned bearings in all — so the arch rotates rather than cracks as it moves. Expansion on a hot day is said to raise the crown of the arch by as much as eighteen centimetres, though that figure is not one any statutory or institutional source states, so it is better given as a measurable rise than as a precise number.

The span is 503 metres, the total length 1,149 metres, with 646 metres of approach spans. The State Heritage Register inventory's 1,670 feet is an outlier against 1,650 feet everywhere else and looks like a transcription slip.

Who designed it, and why that is still disputed

The statutory records split the credit. The State Heritage Register names the designers as J. J. C. Bradfield and R. Freeman with Dorman Long and Co. as builder. The Commonwealth's national heritage assessment is more careful: Bradfield, as the state's chief engineer, prepared the general design and specification and supervised the project, and was largely responsible for bringing the bridge to fruition; Dorman Long won the contract on 24 March 1924 and retained the English consulting engineer Sir Ralph Freeman for the detailed design.

John Job Crew Bradfield, 1867 to 1943, was chief engineer for metropolitan railway construction from 1913, advised on the Sydney Harbour Bridge Act 1922, recommended accepting the Dorman Long tender and, after being retired from the railways in 1930, represented the government with the contractors and oversaw construction. The scheme is his: not only the crossing but the underground railway and the whole north-shore transport plan it was built to serve. The approach road carries his name.

The quarrel began in his lifetime. Freeman published articles in a Sydney newspaper in 1929 describing himself as the designer; the design had been modified after his visit in 1926; Dorman Long threatened to sue over a plaque naming Bradfield as designer; some accounts credit the detailed work to Lawrence Ennis, the contractor's director of construction; the engineer Crawford Munro later wrote that Bradfield did not design the bridge now standing. Bradfield replied that he had originated the cantilever design recommended in 1913. In 2007 the corporate successor to Freeman's firm pressed his claim publicly again. The Australian Dictionary of Biography, which is the soundest authority on the question, declines to adjudicate and says only that the controversy was never finally resolved.

The honest formulation is that Bradfield conceived the scheme, wrote the specification and supervised the work, that Freeman produced the detailed design under contract, and that which of them designed the bridge is a question their contemporaries could not answer either.

Port Jackson, Sydney Harbour and the river

The relationship is the reverse of the way it is usually stated. Port Jackson is not the official name of Sydney Harbour — it is the name of the whole port, defined as all the waters inside a line joining North Head and South Head, and Sydney Harbour is the main central body within it. Middle Harbour, North Harbour, the Lane Cove River and the Parramatta River are the other parts.

So the Parramatta River does not join the harbour at an inland terminus: it is the port's western arm, meeting the rest of it between Greenwich Point and Robinsons Point, well inside the heads. Port Jackson's length is given as 19 kilometres, which is where the figure of twenty kilometres attached to the harbour comes from.

Before the bridge, and the twenty kilometres that were saved rather than travelled

Crossing to the north shore meant a boat — passenger ferries and vehicular punts, one of which ran from a ramp at Dawes Point — or a long drive around the head of the Parramatta River. The overland alternative ran over the Pyrmont, Glebe Island, Iron Cove, Gladesville and Fig Tree bridges, and it is from this that the "five bridges" story comes. The twenty kilometres in circulation is a saving, not a distance: that route took twenty kilometres off the older way round. The journey still took several hours, and no official source gives its length.

Building it, and the date that is almost always wrong

Construction did not begin in 1926. The first sod was turned on the approaches on 28 July 1923; the State Heritage Register puts the first work on the bridge itself in 1924; excavation for the abutments and pylons began in January 1925 and the southern abutment's foundation stone was laid on 26 March 1925. The approach piers were finished in September 1926, and that completion date, read as a start date, is almost certainly the source of the error.

Arch erection began on 26 October 1928, the two halves met on 19 August 1930, and the deck was finished in June 1931.

It opened on Saturday 19 March 1932, and Premier Jack Lang performed the opening — not a royal or vice-regal figure, which is the detail most popular accounts lose; Governor Philip Game inaugurated the bridge's name. Before Lang could cut the ribbon, Captain Francis de Groot of the New Guard rode a borrowed horse out of the crowd and slashed it with a cavalry sword. He was arrested and fined five pounds, and Lang cut a replacement. More than 750,000 people came, and two thousand homing pigeons were released.

The steel, the rivets and the ninety-six locomotives

The steel figure in circulation is wrong in both number and unit. The bridge contains about 52,800 tonnes of steelwork including the approach spans, of which some 39,000 tonnes are in the arch alone. The 58,000 that appears everywhere is the same quantity converted into American short tons, so quoting it as tonnes doubles the error.

On rivets, the sources genuinely disagree and it is worth saying so: the State Heritage Register inventory gives five million, made by McPherson's of Melbourne, while the engineering and library sources give about six million, hand-driven. The register is the statutory record and the others are the better-attested figure, which is an unresolvable pairing rather than a mistake.

The load test is a good story told slightly wrong. Ninety-six steam locomotives were used, over about three weeks in early 1932 — one source puts the start in January and two in February — and they stood end to end across all four railway tracks, not along the roadway.

Sixteen men died. The Commonwealth assessment draws a distinction worth keeping: of those sixteen deaths across the bridge's workplaces, seven were of men employed on the structure itself, the rest in quarries, workshops and on the approaches. "Sixteen died building the bridge" and "sixteen died on the bridge" are not the same claim.

What it cost, and the toll that outlived the debt

The cost cannot be given as one number, because the published figures cover different things. The structure itself came to a little over £4.2 million; with approaches, land resumptions and interest the total is given variously as £6.25 million and as closer to £10 million. Dorman Long's winning tender was £4.5 million against £5.3 million from the other bidder. The State Heritage Register avoids pounds altogether and records a final cost of more than seventy million dollars, about 350 per cent above the original forecast, from a government disclosure of 1988.

The 1932 toll schedule is the most charming document attached to the bridge: sixpence for a car, threepence for a horse and rider, twopence for an unridden horse or a cow, a penny a head for sheep or pigs — and twopence each for the seven circus elephants that crossed. So the claim that the traffic once included horses is quite right, and understates the menagerie.

The debt was cleared in 1988 and the toll was kept. What is structurally true, and will stay true, is that the crossing is tolled southbound only, charged electronically with no booths, at rates that vary by time of day across peak, shoulder and off-peak bands. That changes at the end of the decade: the state government confirmed in December 2025 that two-way tolling will come to the bridge and the harbour tunnel when the Western Harbour Tunnel opens in late 2028.

Traffic is about 160,000 road vehicles a day on the transport authority's own figure, with 1.3 million pedestrians and 400,000 cyclists a year.

The pylons, which hold nothing up

This is confirmed by the statutory record rather than merely suspected. The State Heritage Register states that the abutment towers with their pylons are not a necessary structural feature of the bridge, that they do not support the arch, and that they were built principally to improve its appearance. The Commonwealth assessment adds that Bradfield insisted on them, largely for that reason.

They stand 89 metres above mean sea level, and are concrete faced with granite quarried at Moruya; the first stone was laid in October 1925 and the facing was complete by the end of 1931. The design is generally attributed to Thomas S. Tait of the Scottish practice John Burnet and Partners, though neither statutory record names him.

Never the longest, and what it was instead

The comparison with the Golden Gate is wrong twice over. That bridge is a suspension bridge, so there is nothing to turn on its head; and it opened in 1937, five years after Sydney, so it is not even a benchmark of the period.

The real comparison is the Bayonne Bridge over the Kill Van Kull in New York, which is this bridge's near-twin: a steel through arch, built at the same time, opened in November 1931, four months earlier, and longer in span. How much longer depends on what is measured — about seven metres on the usual figures, or six-tenths of a metre for the arch itself on the Commonwealth assessment — but either way Sydney was never the longest arch span in the world, not even on the day it opened. That near-miss is a better story than the one it is usually given.

By span it is one of the ten longest arch bridges, and no ordinal is worth printing because Chinese arch bridges keep opening above it. The claims that do hold are different ones. It was the widest long-span bridge in the world, at 48.8 metres, until 2012. Engineers Australia's position, adopted in the Commonwealth assessment, is that it is the greatest arch bridge in the world by mass and load capacity — a judgement that explicitly sets span aside, which is why it is the safest superlative to repeat. The Americans' own citation calls it the longest single-arch bridge ever built outside the United States.

Getting onto it

The footway is on the eastern side and the cycleway on the western, with the north-shore railway running between the roadway and the cycleway, so the two are on opposite sides of the bridge and are not shared. The southern approach to the footway is by the Cumberland Street stairs and tunnel, at the edge of The Rocks.

The most useful practical change to the bridge in many years happened at the northern end. The Sydney Harbour Bridge Cycleway Ramp, 170 metres long, opened in January 2026 through Bradfield Park between Alfred Street South and the Lavender Street roundabout, and replaced fifty-five stairs — the transport authority's own description is that it makes riding across the harbour accessible to all. It came with a new cycle route along Alfred Street South, new crossings, and a separation of walkers from riders near Milsons Point station.

There are eight road lanes, numbered from the western side, of which the two easternmost were tram tracks until 1958. Several of the middle lanes reverse direction at peak times, so the configuration is not fixed and no single arrangement should be relied on.

Inside the south-east pylon

The attraction up the pylon is no longer the Pylon Lookout. The BridgeMuseum opened on 27 May 2026 in its place, operated by BridgeClimb Sydney and funded by the Hammon family, who own the south-east pylon; it is self-guided, with interpretation in Mandarin, Korean, French, German and Japanese, and a premium after-hours space called SkyHouse sits 87 metres above sea level. The pylon has been a tourist attraction on and off since Archer Whitford opened it in 1934, closing and reopening several times since. The roughly two hundred steps often quoted belonged to the old lookout, and no step count has been published for the new museum.

Climbing the arch

The operation is BridgeClimb Sydney — not, as it is sometimes rendered, a bridge club, which is where people play cards. It launched on 1 October 1998, founded by Paul Cave after nine years of trying, and Hammons Holdings has held the franchise since June 2018 under a twenty-year contract awarded by state tender, so accounts naming the founders as the operators are out of date.

Climbers are issued with a climb suit and attached to a static safety line; personal items are not taken up. Besides the summit climb there is an interior route through the steel, a full crossing north and back, and Burrawa, an Indigenous-led cultural climb introduced in 2021. Departures run at dawn, through the day, at twilight and at night, so a night climb is a time rather than a separate product.

Heritage, and what is coming

The bridge is item 00781 on the New South Wales State Heritage Register, gazetted on 25 June 1999 as "Sydney Harbour Bridge, approaches and viaducts", with the transport authority as owner. It is also on the Commonwealth National Heritage List as place 105888, found to have outstanding value under five of the nine criteria; the inclusion date of 19 March 2007, the seventy-fifth anniversary of the opening, is not stated in the Commonwealth document consulted and should be treated as unconfirmed. Engineers Australia and the American Society of Civil Engineers have both recognised it as a landmark.

Maintenance is permanent rather than periodic: roughly 485,000 square metres of steelwork are repainted continuously with lead-free paint, in rotating work sites that include the Cumberland Street stairs and tunnel. Work on the Cahill Expressway structure began on the eastern side in October 2026 and is to move across it over several years. Scaffolding, a support structure and sensors have gone up on the Lavender Street heritage rail bridge to monitor its condition.

The one firm date beyond the tunnel opening is further out than most. Improvements planned for Bradfield Park South near the northern pylons are described by the transport authority as partly needed to prepare for maintenance for the bridge's centenary in 2032 — which is the first official signal of centenary works, and the only one. Nothing is announced for the ninety-fifth anniversary on 19 March 2027.