Two conductive parts at different potentials need to be kept apart. You can measure that separation two ways: straight through the air between them, or along the surface of the insulation that connects them. Those are two different numbers, set by two different sets of rules, and a design has to pass both.
Clearance is the air gap. Creepage is the surface path. Treating them as the same value is one of the most common mistakes in low-voltage equipment design, and it usually shows up late, on a board that already exists.
The framework for both is IEC 60664-1, which covers insulation coordination for equipment up to 1000 V AC or 1500 V DC. This page walks through what each distance is, what drives it, the rule that ties them together, and how each one is checked.
Table of Contents
The two distances, side by side
Clearance is the shortest straight-line distance through air between two conductive parts. If an arc were going to jump the gap, this is the path it would take.
Creepage distance is the shortest path along the surface of the solid insulation between the same two parts. This is the path that surface contamination, moisture, and tracking follow over time.
On a flat, bare surface with nothing between the two parts, the two distances can be equal. The moment you add a rib, cut a groove, or drop in a barrier, they split apart. Clearance still cares only about the air; creepage now has to follow the contour of the surface.
What sets the clearance
Clearance is sized against the fast stuff: transient overvoltages. These are the short, high spikes that ride in from switching events and lightning. The gap has to survive the spike without flashing over.
The design number for that spike is the rated impulse withstand voltage, or U_imp. It comes from two things: the nominal system voltage, and the overvoltage category of the equipment. Category tracks how close the equipment sits to the service entrance. Gear at the origin of the installation sees larger spikes than an appliance at the end of a branch circuit, so it needs a larger clearance for the same working voltage.
Two more factors move the number:
Field shape. A uniform field between smooth, rounded electrodes holds off more voltage for the same gap than a distorted field around a sharp edge or a point. Most real hardware has edges and points, so it gets sized on the conservative, non-uniform case.
Altitude. Thinner air breaks down at lower voltage. The base values assume use up to 2000 m. Above that, you apply a correction factor that widens the gap.
Pollution has only a small floor effect on clearance. It matters far more for creepage.
What sets the creepage distance
Creepage is sized against the slow stuff: the long-term working voltage the surface actually sees, day in and day out. Transients barely factor in here. What matters is the steady stress plus what lands on the surface.
Three inputs drive it:
Working voltage. Higher steady voltage, longer surface path.
Pollution degree. This rates the micro-environment around the parts, from a sealed clean space to a surface that stays wet and dirty. A cleaner environment allows a shorter creepage path for the same voltage.
Material group. This ranks the insulating surface by how well it resists tracking, using the comparative tracking index (CTI). The bands are:
- Group I — CTI 600 and above
- Group II — CTI 400 to 599
- Group IIIa — CTI 175 to 399
- Group IIIb — CTI 100 to 174
A surface that resists tracking well (Group I) allows a shorter creepage path than one that tracks easily (Group IIIb). See CTI and material groups for how tracking resistance is measured and why it matters.
Put together: creepage grows with working voltage, grows as the environment gets dirtier, and grows as the surface’s tracking resistance drops.
The rule that links them
Creepage can never come out smaller than the clearance it shares a path with. The surface route between two parts is at least as long as the straight air gap between them.
So if your working voltage, a clean environment, and a good material group hand you a creepage value that is below the required clearance, the clearance sets the floor. You size the creepage up to meet the air gap. In practice this happens most on low-voltage circuits in clean, Group I designs.
Reading the path on real hardware
The shortest surface path is not always the one you’d trace by eye. Ribs, grooves, slots, and barriers all change it, and each one behaves differently.
A rib forces the creepage path up one side and down the other. It adds surface length, so it helps creepage. It does little for clearance unless it also lengthens the air route.
A groove only counts as added creepage if it is wide enough. Below a minimum mouth width, you measure straight across the top of the groove rather than down into it and back out. That minimum width scales with pollution degree: a dirtier environment needs a wider groove before the walls count, because narrow gaps bridge with contamination.
An insulating barrier between the parts sends clearance around the edge and creepage along the face. A cemented joint with no air path removes both, as long as the bond holds.
When you check a board by hand, run a piece of fine wire or the tip of a caliper along the actual shortest surface route, and honor the groove rule as you go. The number you get is the one that counts, not the one printed in the layout.
How each one is verified
Clearance is checked two ways. You can measure the physical air gap directly. Or you can prove it electrically with an impulse voltage withstand test, which applies a 1.2/50 µs waveform at the rated impulse voltage. If the gap holds without breakdown, it passes. In some cases an AC or DC withstand voltage stands in for the impulse test.
Creepage is a dimensional check. There is no separate creepage test. You measure the surface path and compare it to the value required for the working voltage, pollution degree, and material group. The verification is in the ruler and the tracking data, not in a high-voltage shot.
The solid insulation sitting between the parts is a separate question. That gets its own withstand and partial discharge testing, because a surface can pass its creepage dimension while the bulk material still has voids or weak spots.
Clearance vs creepage at a glance
| Property | Clearance | Creepage distance |
|---|---|---|
| What it measures | Shortest gap through air | Shortest path along the surface |
| Main driver | Transient overvoltage (impulse) | Long-term working voltage |
| Key inputs | Impulse voltage, overvoltage category, field shape, altitude | Working voltage, pollution degree, material group |
| Sensitive to dirt | Slightly | Strongly |
| Sensitive to altitude | Yes | No |
| Failure mode | Flashover across air | Tracking along the surface |
| How it’s verified | Physical measurement or impulse withstand test | Physical measurement against tables |
| Floor value | — | Cannot be less than the clearance |
FAQ
Can creepage distance ever be shorter than clearance?
No. The surface path between two parts is always at least as long as the straight air gap. Where the creepage table gives a smaller value, the clearance sets the minimum.
Do clearance and creepage both change with altitude?
Clearance does, because air breaks down more easily at altitude. Creepage does not. It’s a surface path, so thinner air doesn’t shorten or lengthen it.
Which drives creepage more, higher voltage or a dirtier environment?
Both push it up, and they stack. A modest voltage in a Pollution Degree 3 setting can need more creepage than a higher voltage in a sealed, clean enclosure.
Does a conformal coating count as insulation for creepage?
It can, under specific conditions on how the coating is applied and qualified. A coating that fully and reliably covers the surface can let the design be treated as a lower pollution degree, which shortens the required creepage.
Is there a single test that proves both distances at once?
No. An impulse or withstand test proves the clearance and the solid insulation. Creepage is confirmed by measuring the surface path against the required value.