The B-29 was conceived before the United States entered the war, when planners feared they might have to strike distant enemies from bases far away. The specification demanded range, altitude and bomb load far beyond the B-17 — and pressurisation so crews could operate for hours above the weather and most defences. Meeting it required a new wing, new engines, new turrets and a new fire-control system all at once, and the aircraft was ordered into mass production “off the drawing board” before testing was complete.

The gamble worked, but only just. The R-3350 engines, the guns and the pressurisation all needed fixing in the field, and the “Battle of Kansas” in early 1944 saw crews rebuilding finished bombers by hand outdoors in winter to get them combat-ready. For all its cost and trouble, the Superfortress gave the United States a strategic reach no other air force could match — and it shaped the post-war bombers and tankers that followed.

Most of the B-29’s defensive guns were not held by the men who fired them. Gunners sat at sighting stations — in the nose, an upper “barber-chair” position, two side blisters and the tail — and the General Electric central fire-control system slaved the aircraft’s four remote turrets to their sights. An analogue computer corrected each solution for gravity and air resistance, for the target’s movement, and for the parallax between sight and turret. According to the Smithsonian, the system delivered accuracy at far greater range than hand-aimed guns, and a gunner could hand control of a turret to another station — even automatically if he was incapacitated.

The R-3350 was rushed into service and never fully shed its reputation for catching fire. Its rear cylinders ran too hot, and flammable magnesium in the crankcase and other parts meant that once an engine fire started it could burn through the wing structure before the crew could react. The flaw killed Boeing test pilot Eddie Allen and his crew when the second XB-29 came down in Seattle on 18 February 1943, and caused losses throughout the war. Revised cowl-flap procedures, better baffling and constant maintenance tamed the problem, but engine reliability remained a worry for as long as the type flew.

The Superfortress was extraordinarily expensive to develop and build. The programme is widely estimated at around $3 billion in wartime dollars — more than the roughly $1.9–2 billion spent on the Manhattan Project — which made it the costliest US weapons programme of the Second World War, more expensive than the atomic bomb it was built to carry. Unit prices varied by model and year: one museum cites a 1944 figure near $605,000, with averages across the run put closer to $900,000. These totals are period estimates repeated across many sources rather than audited final accounts, and should be read as order-of-magnitude figures.