What is actually happening

The glass is colder than the air. Not because anything is wrong with it — because it has been radiating heat to the sky all night.

Any surface with a clear view of the open sky loses heat by infrared radiation to space, which is effectively at −270°C. On a clear, still night that loss outpaces what the surrounding air can supply, and the surface ends up several degrees colder than the air around it. Grass does it. Car roofs do it. Glass does it particularly well.

When the pane falls below the dew point of the outdoor air, the air gives up water onto it. That is dew, forming on a window instead of a lawn.

Why it means the glazing is working

This is the part that surprises people, and it is worth stating plainly.

For the outer pane to get that cold, it has to be thermally disconnected from the warm room behind it. That is precisely the job of a modern sealed unit: a low-emissivity coating and an insulating gas fill mean very little heat leaks out to warm the outer surface.

Old single glazing almost never shows this, because heat pouring out of the house keeps the outside of the glass warm. So the sequence is genuinely backwards from intuition:

What outside condensation tells you about the glazing.
GlazingOuter pane at dawnOutside condensation
SingleWarmed by heat escaping the roomRare
Older doubleSlightly warmed from insideOccasional
Modern low-E unitTracks the outdoor air, then radiates below itCommon on clear nights

The three conditions

It needs all three at once, which is why it is not an every-night event:

  1. A humid night — the dew point has to be close to the air temperature. A large dew point spread means the glass would have to fall further than radiative cooling can take it.
  2. A clear sky — cloud radiates heat back down and largely cancels the effect. This is why it happens on the nights that look most pleasant.
  3. Still air — wind mixes warmer air across the surface and keeps it near the air temperature.

Take away any one and the glass stays dry. That combination is most common in late spring and summer, when nights are humid but the ground has cooled.

When it clears

Usually within an hour or two of the sun reaching the glass. Nothing needs doing — as the pane warms past the dew point the water evaporates back into the air it came from.

If it is still there at midday, the sky has stayed overcast and the air is close to saturation, in which case the whole outdoors is damp and the window is the least of it.

The same effect, elsewhere

Once you recognise radiative cooling, a lot of unrelated-looking things turn out to be this:

All the same mechanism: a surface that got colder than the air, and air that had water to give.