If you’re thinking that you need to retract the gear to achieve the minimum climb performance to continue the climb after takeoff, you’ve already done it wrong! Do not be looking at instruments to observe the first moment when you can retract the wheels, fly the plane first. When the plane is climbing steadily, you can select the gear up. The means of determining the steady climb is not critical, the climb is! It’s a discipline thing.
My flying boat has more drag with the wheels up than down. Therefore, when flying over land, I leave the wheels down. When flying extended over water, I retract the wheels. There is no one size fits all for this, other than think about what you’re doing. Before every RG landing state out loud, the position of the wheels, and the surface you intend to land on. Commonly, it will be “wheels are down for landing on land”. Say it as you check, then say it as you double check short final – there is zero excuse to not speak it out loud.
If you choose to ditch or land onto a rough surface, you will probably say “wheels are up for ditching” or “wheels are up for a forced landing” and that’s as it should be. Other possibilities could be: “Wheels are retracted for landing on the water” (amphibian) or “wheels are retracted for landing on snow” (wheels/ski plane). DO not rely on a landing gear position warning system, look and say the words out loud. Wheelplane RG landing gear position warning systems are typically activated by a combination of the gear being up, and the throttle being closed. So if you power your way down final, and never close the throttle, you’d never get a warning horn. Many planes have been bellied in, because the pilot did not hear the horn, even though the wheels were up – ’cause the throttle was only closed after the slide down the runway.
As the Transport Canada Test Pilot and I joked when I rode right seat for his testing (TC requires a right seat PIC to be their “insurance”) – “silence is not a warning”. (I reminded him that it is if it’s your wife being silent, but that’s another topic). Do not rely on a warning system, thinking that no warning is situation normal. You are the pilot, you are responsible. You determine and confirm the position of the gear, and speak it out loud as you do.
When I train amphibians, I pre brief that I will remind the candidate once, and after that, if the landing gear position is not visually confirmed and spoken twice before each landing, I will call a go around. I call it late, and I call it loud!
A puddle jumper VSI will indicate a steady descent or climb, yes there is a lag, but when established in the climb it will indicate correctly. It is fair to say it is a universal SOP to call out positive rate (when confirmed by the VSI), speed check before retraction.
Performance B MEP with RG need to have the gear retracted to meet the net take off flight path obstacle clearance path.
I don’t see that another reason has been mentioned before: If you do not fly our small planes but heavy metal, the Cockpit can do a quite significant “climb” without the main gears living ground. So positive rate of climb ensures that you actually climb and not just rotate.
When I had the Aerostar I got in the habit of pulling gear up as quick as possible for one reason: The 601P has a mushy slow accelerating stage between takeoff speed and blue line (it’s one of the quirks of the fast wing the 601P has – it’s just not designed to climb very well below a certain speed). After that it climbs really well, but not before that. So you wanted to transition that “dead man” zone as fast as you could and eliminate drag.
Also, I do think there’s no need to pull up gear fast in a single in case of an emergency, but I think for twins there’s no benefit in delaying it. Once you’re flying in a twin, if you’re having an engine failure, you’re most likely continuing. Why carry extra drag?
There is no single answer when to set the gear up in GA.
On big jets its somewhat simpler. Positive rate – observed on VSI (inertial), altitude rising – gear up. That’s it. You are always above V1 when airborne.
On GA it should be adjusted a bit according to actual conditions. Short runway – obstacles- go minded – gear up asap – lets minimize drag.
3 km runway – I would delay 10-15 secs to give myself a chance to abort in case of eng failure even on a twin.
Raven wrote:
Short runway – obstacles- go minded – gear up asap – lets minimize drag.
Again, it depends on the type. In Cessna RGs you leave the gear down until obstacles are cleared as you lose some climb performance during the retract process.
172driver wrote:
In Cessna RGs you leave the gear down until obstacles are cleared as you lose some climb performance during the retract process.
Of course. There are also aircraft-specific properties.
On some other types you have to be careful – not to exceed max retraction speed when deciding to continue with the gear extended which is a factor pushing you to retract quickly or to control your pitch in order to keep proper speed.
172driver wrote:
In Cessna RGs you leave the gear down until obstacles are cleared as you lose some climb performance during the retract process.
But you lose 40cm obstacle clearance with gear down no?
Probably not worth it vs 10s of retraction drag 
Raven wrote:
I would delay 10-15 secs to give myself a chance to abort in case of eng failure even on a twin.
On my checklist the gear up is done as soon as the remaining runway is not long enough for a (forced) landing.
With the Cessna Skymaster one must not retract the landing gear during an EFATO, at least not while close to the ground:
b. A note in the section titled ENGINE-OUT ON TAKE-OFF stated: ’Airplane drag with the landing gear doors
opened and the gear partially extended is greater than the drag with the landing gear fully extended. Corresponding
rate-of-climb penalties are -240 ft/min and -110 ft/min respectively.
[…]
An additional climb penalty would have arisen as the landing gear doors opened during the retraction cycle.
[…]
Factors
2 The pilot retracted the landing gear
