The launch pad at the Peenemünde research station, deep in the pine forests of the island of Usedom on the remote Baltic coast. All the test flights of the A4 were initiated from this site.

By then, the research group was fast outgrowing the facilities at Kummersdorf. The A2s were actually launched from the island of Borkum, off the mouth of the River Ems in the North Sea. In early 1936, the HWA and the RLM joined forces to purchase a large area of suitable land, an isolated peninsula around the village of Peenemünde on the island of Usedom, off the Baltic coast close to the present-day border with Poland. They also bought the adjacent Greifswalder Oie, and it was to this location that both moved their rocket development programmes. The Army were located to the east of the site, in an area known as Heeresversuchsanstalt-Peenemünde (HVP), and referred to as Electromechanische Werke (EMW) as a cover, where Dornberger was appointed Head of Weapon Test Section 11, with von Braun as his Technical Director. It was here that the first successful ballistic missiles the world had ever seen were developed and tested, and the name Peenemünde soon took on a new significance.

After the failure of the A3 to live up to expectations, development work on the A4, which was always intended to be a military rocket, was halted. Instead, von Braun turned to the design of another research rocket, the A5, somewhat larger than the A3 but with the same motor. The main difference between these and the earlier design was to be in the profile of its flight path, because now the requirement was not simply to send a rocket straight up into the atmosphere, but to launch it at a terrestrial target hundreds of kilometres away, and for that, a sophisticated guidance package was required. In the case of a small device like the V1, basic direction was simply a matter of aligning the launch ramp with the azimuth of the target, and then relying on a gyrocompass to apply small corrections. But steering the ballistic missile to its target would be quite another matter, since it would have to be launched vertically, and then tipped over in the precise direction of the target to an angle of 41 degrees from the horizontal and maintained there. Range was determined by the length of the burn, and that meant that propellant cut-off had to be precise and instantaneous. Cut-off was actuated initially by a radio signal from the ground, and was the only external factor applied after the launch sequence had been initiated (and this, too, being later automated). Without going too far into the intricacies of the matter, to achieve a proper degree of directional stability in a ballistic missile it is necessary to be able to control its movement in three axes: pitch (to achieve and maintain the proper angle of climb); yaw (side-to-side movement, to correct the heading); and spin, which is a natural tendency of a cylindrical body in motion, but which makes controlling pitch and yaw by means of rudders almost impossible, and which must be damped out. To make matters worse, the characteristics of the missile – in particular the all-important centre of gravity – change as its fuel is consumed, and its flight characteristics change no less drastically as it climbs into and through the upper atmosphere and then descends again on its parabolic course.