Power was provided by four Rolls-Royce Proteus marine turboshaft engines each driving its own lift fan and pylon-mounted steerable propulsion propeller.

The SR.N4 was the largest hovercraft then built, designed to carry 254 passengers in two cabins besides a four-lane automobile bay which held up to 30 cars.

In response to operator demands, stretched versions of the SR.N4 were developed, culminating in the Mk.III variant, which had almost double the capacity for carrying both cars and passengers as the Mk.I.

In its place, a new proposed hovercraft, which was referred to as the SR.N4 as well, was considerably larger and heavier, and powered by three pairs of Rolls-Royce Proteus marinised gas turbine engine.

[7][8] This approach is credited with having been less expensive and having resulted in a more commercially viable hovercraft than would have been if it had been constructed as per the earlier incarnation of the SR.N4 and then required to perform modifications to improve the capabilities of aspects such as the skirt, which had been considerably advanced by development of the SR.N5 and SR.N6.

[9] Specific improvements included the adoption of triangular rubber 'finger'-like attachments to the curtain which provided for a better seal between the hovercraft and the water's surface as well as being cheaper and easier to maintain than previous configurations.

[11] In November 1965, Frank Cousins, the Minister of Technology, announced that British Rail would participate in the development of the SR.N4 and would be a customer for the type.

[20] The air intakes had to be substantially modified in order to minimise salt ingestion, and a revised skirt system was also developed.

[22] Six independent electrohydraulic systems, each driven from the main gearboxes, powered the movement of the fins and pylons, while a further four units were used to actuate the variable-pitch propellers.

[26] It housed a crew of three, comprising a captain, first officer/flight engineer, and a second officer/navigator; the main role of the second officer was to avoid collisions, primarily using a Decca 629 radar to do so.

[30] The stern of the craft featured a sizeable set of doors for the loading and unloading of vehicles onto the car deck as well as all four of the exhausts for the Proteus engines.

[31] Upon completion of the prototype SR.N4, Charles Anthony Brindle, the Managing Director of British Rail Hovercraft, was responsible for establishing the first cross-Channel route for scheduled services by the type.

[32] In October 1966, Brindle and several engineers surveyed several potential sites on both the British and French sides of the English Channel using an SR.N6 to determine their suitability for the hovercraft service, which had been given the name Seaspeed.

[33][34] It was soon determined that Dover and Boulogne would be the preferable option for the maiden route, a decision that had been motivated in part by the local Chamber of Commerce having agreed to build a suitable hoverport to readily facilitate such operations.

Brindle was aware that BHC would not be paid until the SR.N4 was in service, and thus there was a considerable emphasis placed on getting the craft ready for commercial use right at the point of delivery.

The fastest ever crossing of the English Channel by a commercial car-carrying hovercraft was 22 minutes, achieved on 14 September 1995 by the SR.N4 Mk.III GH-2007 Princess Anne on its 10:00 a.m.

[40] To maintain speed, the engines were upgraded to four 3,800 shaft horsepower (2,834 kW) models, which were fitted with four 21 ft (6.4 m) diameter steerable Dowty Rotol propellers.

The work cost around £5 million for each craft, and they were designated Mark IIIs; the improvements allowed them to operate in seas up to 11 ft 6 in (3.5 m) high and with 57.5-mile-per-hour (92.5 km/h) winds.

[44] However, in March 1985, 4 passengers lost their lives when GH-2006 Princess Margaret crashed into a pier at the entrance of Dover Harbour from Calais in force 6 to 7 heavy seas.

A minesweeping version of the SR.N4 was thought to be capable of carrying the same equipment as a Hunt-class mine countermeasures vessel, while being cheaper to purchase, although more expensive to operate.