Two questions that get answered by one number
Every conductor selection has to satisfy two independent constraints, and people routinely solve one and assume the other came along for the ride.
The first is ampacity: can this conductor carry this current continuously without its insulation exceeding the temperature it is rated for. That is a safety question, it is answered by tables in the adopted code, and it is heavily modified by the installation conditions. The second is voltage drop: will enough voltage still be there at the far end for the equipment to work properly. That is a performance question, it is answered by arithmetic, and the installation conditions barely touch it.
On short runs, ampacity is the binding constraint and voltage drop is irrelevant. Twenty feet of 12 AWG on a 20 amp circuit drops well under a volt. On long runs it flips, sometimes dramatically. A 20 amp load at 300 feet on 120 volts needs something in the 4 AWG range to hold 3 percent, while ampacity alone would have been satisfied by 12 AWG. This page solves the voltage drop side and hands you a floor. The ampacity side has to be checked against the table your jurisdiction actually enforces, with the correction factors for your actual installation applied.
Working the formula backwards
The forward calculation is drop equals multiplier times K times amps times feet divided by circular mils. Rearranged for the area you need:
Circular mils = (multiplier × K × amps × one-way feet) ÷ allowed volts of drop
Allowed volts is your target percentage times the supply voltage. Three percent of 240 volts is 7.2 volts. The result is a raw area, almost never a size that exists, so you round up to the next standard conductor. That rounding is why the candidate table above shows the sizes either side: sometimes you land a hair over a size boundary and going down one gets you 3.1 percent instead of 2.4, which is a judgement call rather than a violation of anything.
Two inputs deserve care. The first is the load current: use what the circuit will actually draw, not the breaker rating. Sizing a conductor for voltage drop at 50 amps when the load is 28 amps buys copper you do not need. The second is distance: measure the route the cable will take, including the vertical runs and the detour around the obstruction, not the straight line on the plan. People underestimate this by 20 percent as a matter of routine.
Continuous loads and the 125 percent figure
A load that runs for three hours or more at a stretch is treated differently. The widely applied convention is that the conductor and the overcurrent device are sized for 125 percent of the continuous load, which is the same thing as saying the device is not loaded beyond 80 percent of its rating. Electric heat, EV charging, lighting in a commercial space and long-running shop equipment all fall in this category. A 40 amp continuous load is sized as 50 amps.
That 125 percent applies to the ampacity and breaker side of the problem. For voltage drop the honest figure to use is the actual current, because that is what actually flows. The checkbox on this page shows you both so the two numbers do not get confused with each other when you are talking to a supplier or an inspector.
What to hand to the electrician
If you came here to arrive at a conversation better prepared, the useful things to bring are the ones a calculator cannot infer: the actual load in amps and whether it is continuous, the measured route length, the system voltage and phase, where the run passes through and what the ambient temperature is along it, how many other current-carrying conductors will share the raceway, what the terminations at both ends are rated for, and whether the wiring method is conduit, direct burial, cable in framing or something else. With those in hand the sizing conversation takes five minutes. The number this page produced is one line of that list, and it is the line that is the same everywhere. The rest of it is why the person doing the work needs to see the site.
Questions people ask
What size wire do I need for a 50 amp circuit?
That question cannot be answered from the breaker rating alone, and answers that come back instantly are skipping the parts that matter. What determines it: the actual load and whether it is continuous, the insulation type of the conductor, the temperature rating of the terminals at the breaker and at the equipment, the ambient temperature along the route, how many current-carrying conductors share the raceway or cable, the wiring method, the length of the run, and the code edition your jurisdiction adopted with its local amendments. Two identical 50 amp circuits can legitimately need different conductors in different buildings. Use this page for the voltage drop half of the answer, and get the ampacity half from someone who can see the installation.
Should I size for the breaker or for the load?
Ampacity is coordinated with the overcurrent device, so on that side the conductor and breaker are sized together and the conductor is generally not smaller than the device it is protected by. Voltage drop is caused by the current that actually flows, so on that side you use the real load. A 50 amp breaker feeding a 22 amp load produces the voltage drop of a 22 amp load, not of a 50 amp one, and sizing the run for 50 amps of drop over a long distance can add a size or two of copper for nothing. Where the two disagree, the larger conductor wins, because the ampacity requirement is a safety requirement and voltage drop is a performance target.
Is aluminium worth it for a long feeder?
Frequently, on cost, and that is why utility and service-entrance conductors are so often aluminium. You will typically need to go up about two sizes to match copper on both ampacity and drop, and at large sizes the material saving still comes out well ahead. The trade-offs are real, though: the terminations at both ends have to be rated for aluminium, they need to be prepared and torqued the way the connector manufacturer specifies, the conductor is bulkier so the conduit fill and bending radius change, and some equipment lugs simply will not accept it. This is a decision to make with the person installing it rather than in a shopping cart.
Does conduit fill or the number of conductors change the answer here?
Not for voltage drop, which is what this page calculates. It changes the ampacity side substantially. Bundling several current-carrying conductors together means each one has less ability to shed heat, and the commonly applied adjustment reduces the allowable ampacity progressively as the count rises, with a further correction if the ambient temperature is above the table baseline. Those two corrections stack multiplicatively and can take a conductor well below its table figure. Neutrals that carry current in unbalanced or harmonic-rich systems count toward that total, which surprises people. It is one more reason the ampacity half of the decision needs a person, not a page.