I'd like to make a controversial claim: Drogues should be sized based on the surface area of the rocket, not the weight. My LPR BT-70 rocket comes down far more slowly drogueless than my 54mm FG did when I had an accidental drogueless. But of course they do. There's a similar amount of drag, and much less mass. A 4lb minimum diameter rocket will require a different sized drogue than a 4lb big fat light rocket. I see the drogue as having three purposes: Prevent the collection of parts from coming down nose-down streamlined. Provide a stabilized "Flying-V" platform to allow for an undisturbed main deployment Slow the assembly down to a speed where the main can deploy safely Let's start with two parts of the rocket falling separately. The rate at which they will fall (terminal velocity) is increased by weight, and decreased by area. Terminal velocity is determined by ballistic coefficient (mass/area), or sectional density Let's assume that the fin can has a higher ballistic coefficient than the payload, which is likely - motor casings are heavy and parachutes are light. The fin can will fall faster than the payload, and since they are attached by the shock cord that will pull the forward end of the fin can up. Not bad. Fin can comes down, since the payload falls more slowly it will pull on the shock cord and pull the nose of the fin can up. Now the fin can is falling tail first, which is unstable, so it will then turn nose down again. Pretty flat and high drag. Now let's look at the unpleasant, but unlikely case: Payload has higher sectional density/ballistic coefficient than the fin can. Now the payload pulls the nose of the fin can down, and they both come down streamlined: Not good. This could occur if the break is too far up the rocket, and the fin can section is mostly air, while the payload section is a tightly-packed parachute So it looks like the important factor here is not the weight of the rocket, but the ballistic coefficient. If both sides have about the same ballistic coefficient, or the fin can has slightly more, it will come in stable drogueless: Now let's throw the drogue out on our stable configuration: Same amount of mass pulling down, but now we've got more area pulling up. So the whole assembly will now fall more slowly. Since the whole assembly is falling more slowly, the fin can and payload are now below terminal velocity, so they will start falling faster/drooping and pulling on the parachute. This decreases their surface area, causing speed to pick up. If the system is designed properly, it will quickly stabilize at speed where the drogue is taking up some of the weight, but the fin can and payload are still "flying" and making significant drag. That balance point is the V we want. So here's my controversial claim: For a perfect flying V, the drogue is creating about one third to one half of the drag, leaving the fin can/payload to "fly" and support the other half of their weight. All that matters in sizing the drogue is the side-on surface area presented by the rocket. If you have a rocket that's 100" tall and 3" wide, that's 300 in sq of area, and a good estimate for drogue size would be 100-150 sq in, or a 10-12" chute. That will give a good balance between flying the body and keeping the line taut and give us the "flying V". (Here I am making the assumption that the Cd of a horizontal cylinder and a flat parachute are about the same - about 0.8. If you're using a high-Cd drogue like a Fruity or Rocketman, an even smaller drogue works. If you're using a streamer, Cd of 0.2 or so, you can get away with much more area) If that 3"x100" rocket is heavier, it will come down faster, and if it is lighter, it will come down more slowly, but the angle of the V is determined not by the weight of the rocket, but by the ratio between the surface area of the rocket and the surface area of the parachute Second controversial claim: At the weights/densities of the average HPR rocket and the strength of HPR parachutes, any rocket that's coming down slowly enough to form a good flying V will be slow enough to allow for a safe deployment of the main. At most, a slider ring will be required.