In 1961, before the Saturn V had come into being even on paper, the question was “how will we land on the moon?” The general assumption was that we’d use the “direct” approach. In other words, the manned vehicle would land directly on the lunar surface. No dedicated Lunar Excursion Module, no lunar orbit rendezvous. The Apollo capsule itself would form part of the luanr lander, and would have to be launched from the luanr surface. This meant that the lunar lander would be, compared to the LEM, fairly enormous. Which meant that the launcher would need to be either fairly enourmous… or there would need to be multiple launchers.
In October of 1961, NASA-Lewis produced a presentation showing several options. The obvious approach was the use of the Nova class rocket. This would be a large but conventional rocket with eight F-1 engines on the first of four stages. While shorter than the later Saturn V, the notional Nova described would have a gross weight close to 50% greater, at 9.6 million pounds.
An alternative approach would use the Saturn C-3. The C-3 concept used two F-1 engines (later versions used 3) on the first stage; twice as capable as the Saturn I, it still fell far short of being able to do the job on its own. So two Saturn C-3s would be needed per mission. The first C-3 would launch a liquid oxygen “tanker” into Earth orbit; the second would launch the actual lunar landing spacecraft. After topping off the LOX tanks of the lunar lander & upper stage, the spacecraft would go on about the mission. The total launch weight would be substantially reduced to 4.8 million pounds.
A third approach was proposed. A single Saturn C-3 would be used. But this time, the upper stages would be equipped not with A-3 (i.e. early RL-10) rocket engines, but a single nuclear thermal rocket. Equipped with a large liquid hydrogen propellant tank, this single booster would be able to carry out the whole mission. Launch mass was again substantially reduced, this time to 2.4 million pounds.
