The short answer
| Limit | Applies to | Status in the NEC |
|---|---|---|
| 3% | A branch circuit on its own | Informational Note to 210.19(A): explanatory, not enforceable |
| 3% | A feeder on its own | Informational Note to 215.2(A): explanatory, not enforceable |
| 5% | Feeder plus branch circuit combined | Same Informational Notes: explanatory, not enforceable |
So: widely used, sensible, and not a code requirement in the NEC itself. That distinction matters, because it means neither "the code says 3%" nor "voltage drop doesn’t matter, it’s just a note" is a correct summary.
What the NEC actually says
The figures live in Informational Notes attached to the branch-circuit and feeder articles. NEC 90.5(C) is explicit about what that means: Informational Notes are explanatory material and are not enforceable as requirements of the code.
They are the code making a recommendation rather than a rule. The NEC’s own stated purpose is the practical safeguarding of persons and property from hazards arising from the use of electricity, and excessive voltage drop is mostly a performance and efficiency problem rather than a safety hazard, which is a reasonable explanation for why it was never written as a requirement.
When it becomes mandatory anyway
In a great many real projects a voltage drop limit is binding, just not through the NEC’s general articles. The usual routes:
- Local amendment. Jurisdictions adopt the NEC with changes, and some promote the voltage drop notes into requirements. This is the most direct route and the one to check first.
- Energy codes. Commercial energy standards such as ASHRAE 90.1 and the codes derived from them contain voltage drop provisions for feeders and branch circuits as an efficiency measure. Where an energy code is adopted, those provisions are enforceable.
- Utility requirements. Service and metering rules often set limits on the customer side.
- Equipment specifications. A motor, drive, or piece of electronics with a stated minimum input voltage sets a hard limit of its own, and often a tighter one than 3%.
- Project specification. On engineered work, the design documents set the number and it is contractual.
Choosing a target
Three questions get you to a defensible figure.
What does the equipment need?
Start at the far end. If the data sheet says the device operates down to 108 V on a 120 V nominal supply, the equipment itself has given you a 10% budget, but that budget has to absorb the utility’s own variation as well as your wiring. Utility supply is typically allowed to sit a few percent either side of nominal, so the wiring gets what is left.
How much of the budget does this run get?
A load is usually at the end of a chain: service, feeder, branch circuit. The 5% combined figure exists because the drops add. If the feeder already spends 2%, the branch circuit has 3% left, not 5%.
What will change later?
Loads grow. Circuits get extended. A design that lands exactly on the limit today fails the first time someone adds an outlet. Leaving 20% of the budget unspent costs very little at the design stage and buys a great deal of room later.
When to go tighter
| Situation | Why |
|---|---|
| Low-voltage DC | Small supply means a small absolute budget, and the loss is energy you generated or stored and then threw away |
| Motor starting | Inrush is several times running current, so starting drop is several times the steady-state figure |
| Continuous heavy loads | Power lost is I²R for as long as the load runs; on a large feeder that is a real operating cost |
| Sensitive electronics | Switch-mode supplies draw more current as input voltage falls, which makes the drop worse rather than better |
| Long feeders with future load | Whatever margin you leave now is what a later addition gets to spend |
When looser is defensible
Not every circuit deserves the same care. Where the load is intermittent, resistive and tolerant, spending money on conductor to chase the last percent is not engineering, it is superstition.
- Short-duty resistive loads. A heater that runs a few minutes at a time and does not care about a 4% drop beyond taking slightly longer.
- Loads with their own regulation. An LED driver holding constant output across a wide input range is indifferent until it falls out of that range, but note that when it does, the failure is sudden.
- Lightly loaded circuits. A calculation at full rated current on a circuit that will realistically see a quarter of it is answering a question you do not have. Calculate at the actual expected load, and then confirm the assumption is still true when someone adds to the circuit.
Questions
What is an acceptable voltage drop percentage?
3% on a branch circuit and 5% for feeder plus branch circuit combined are the figures most designers work to. In the NEC these appear in Informational Notes, which under NEC 90.5(C) are explanatory and not enforceable requirements. Treat them as sound engineering targets rather than as code limits, and check what actually applies to your job: local amendments, energy codes, utility rules and equipment specifications can all impose real limits.
Is 3% voltage drop a code requirement?
Not in the NEC itself. The 3% branch-circuit figure appears in an Informational Note to 210.19(A), and NEC 90.5(C) states that Informational Notes are explanatory material and are not enforceable as requirements of the code. That said, a number of jurisdictions adopt the NEC with local amendments that make it mandatory, and commercial energy codes such as ASHRAE 90.1 impose equivalent limits, so in many real projects a 3% limit is binding through some other route.
Is 5% voltage drop too much?
It depends entirely on the load. 5% is the usual outer bound for feeder plus branch circuit combined, and most general-purpose equipment tolerates it without complaint. But a motor at 5% low voltage produces about 10% less torque, and a starting motor sees far worse because inrush current multiplies the drop. Sensitive electronics, long low-voltage DC runs and equipment with a stated minimum input voltage all need tighter figures.
What voltage drop is acceptable for a 12 V system?
There is no code figure for low-voltage DC in premises wiring, so practice is set by what the equipment needs. 3% is a common working target for a main run and many solar installers design to 2% or tighter between array and charge controller, because that loss is energy that was paid for and never recovered. The real constraint is usually the minimum input voltage on the device at the far end.
Does voltage drop matter more for motors?
Yes, in two ways. Motor torque falls roughly with the square of applied voltage, so a 5% drop costs about 10% of available torque. And to deliver the same mechanical output at reduced voltage the motor draws more current, which increases the drop further and heats the windings. Starting is worst: inrush current is several times running current, so the drop during a start is several times the steady-state figure, which is why a motor that runs happily can still fail to start.
Set your own limit on the calculator. 3% and 5% are one click away, and so is anything else.