Perseids 2026: a decision-first viewing guide

The 2026 maximum falls on Thursday 13 August 2026 at 02:00 UTC, 8 hours after new Moon. The Moon is 0% lit. That removes the one factor you can do nothing about, and it changes what the rest of the decision looks like.

Why this year is different

In most years the honest advice about the Perseids is hedged, because the Moon usually takes a large bite out of the peak. Not this one. New Moon falls on 12 August 2026 at 17:37 UTC, and the shower's maximum arrives 8 hours later. There is effectively no Moon in the sky at all on the peak night.

Across 2026 to 2030 no other Perseid maximum comes close. The next darkest is 2029, with the Moon 10% lit; 2027 is a write-off at 87%. If you are going to make one deliberate effort for this shower in the second half of the decade, it is this one.

Worth being precise about what "no Moon" buys you. Moonlight does not reduce the number of meteors arriving; it raises the sky background so the faint ones become invisible. The Perseids have a population index of 2.2, meaning each magnitude of sky brightness divides your count by 2.2. A gibbous Moon costs roughly two magnitudes, so a moonlit Perseid peak runs at something like 21% of a moonless one. In 2026 you keep all of it.

The numbers this shower actually runs on

Zenithal hourly rate

100 — second only to the Geminids at 150

Population index

2.2 — bright-rich, so it tolerates an imperfect sky

Speed

58.8 km/s — fast, so meteors are brief, bright and often leave trains

Radiant declination

+57.88° — high in the north, absent from the far south

Activity window

17 July – 24 August — 39 days

Parent body

Comet 109P/Swift-Tuttle

The zenithal hourly rate is a ceiling, not a forecast: it is what a single observer would count under a magnitude 6.5 sky with the radiant directly overhead. Two corrections cut it down before you ever get outside. The first is geometric and depends entirely on where you are.

Who this actually works for

A radiant's highest possible altitude is 90° minus the difference between your latitude and its declination, and the observed rate scales with the sine of that altitude. For a radiant at +57.88° that produces a very lopsided map:

South of 32.1°S the Perseid radiant does not rise at all, on any night of the year. Sydney, Cape Town, Auckland and Buenos Aires are all inside that band, and no dark site, clear sky or moonless night changes it — the modelled rate there is exactly zero rather than small. The southern hemisphere's equivalent event is the eta-Aquariids in May, whose radiant sits on the celestial equator.

Which night, and why the answer depends on your longitude

The maximum is an instant, not a night: 02:00 UTC on 13 August 2026. Where that lands in your own darkness decides which night to pick.

• Europe and Africa get it in the small hours of the night of 12–13 August — the best possible placement, since the radiant is also highest before dawn.
• The Americas get it during the evening of 12 August, when the radiant is still climbing. The night of 12–13 is still the one to take, with the strongest hours after local midnight.
• East and South-East Asia get it in daylight. Both adjacent nights — 12–13 and 13–14 August — sit within hours of the maximum, and with a 39-day activity window the difference between them is small compared with whatever the cloud does.

In every case the useful hours are the ones closest to dawn, because that is when the radiant is highest, not because of anything about the stream itself.

What can still ruin it

One thing, mostly: cloud. With the Moon removed, cloud becomes the entire remaining risk, and it is the factor that cannot be known in advance — numerical weather forecasts are useful inside about a day and degrade quickly beyond that. Any article published in July telling you that 12 August will be clear is guessing.

What helps is that the shower is active from 17 July – 24 August. If the peak night is overcast where you are, the nights either side still produce well, and the Moon stays out of the way for several of them. That resilience is the practical difference between the Perseids and a narrow shower like the Quadrantids, where a cloudy peak ends the season.

The second risk is self-inflicted: going out for twenty minutes, checking a phone twice, and concluding the shower was overhyped. Rod-cell dark adaptation takes twenty to forty minutes and a bright screen resets much of it. Budget an hour outside minimum.

Is the drive out of town worth it this year?

More than usual, and for a specific reason. In a moonlit year, escaping the city removes one of two bright sources and leaves the other, so the payoff is muted. In 2026 the Moon is already gone, so every magnitude of sky you gain by driving converts directly into a factor of 2.2 on the count. An hour's drive typically buys one and a half to two magnitudes — call it a factor of 4.0 to 4.8.

Set against that: two hours of your night, and the dark adaptation you lose to oncoming headlights, which you will have to rebuild after you arrive. The trade is usually worth it on a clear moonless peak night and rarely worth it otherwise. Decide it against your own forecast, not against a general rule — and if you do drive, arrive early enough to spend the first half hour adapting rather than looking.

What you should realistically expect

From a mid-northern latitude, under a genuinely dark sky, near the peak, with no Moon: several tens of meteors an hour in the hours before dawn, in clumps rather than evenly. From a suburban garden the same night: perhaps a third of that, still with a steady trickle of bright ones, because at r = 2.2 the Perseids keep a decent share of bright meteors even when the sky is compromised.

A rate model built for a single observer with an unobstructed view will overshoot what you count by 30–50%, because you are one person looking at part of the sky and occasionally blinking. That is normal and it is stated plainly on the methodology page rather than hidden in the small print.

If you miss it

The Southern delta-Aquariids overlap the Perseids almost entirely (12 July – 23 August) at a zenithal rate of 25. They are fainter-skewed at r = 3.2, so they need a properly dark sky, but from southern latitudes they are the better of the two by a wide margin. Later in the year the Geminids at a rate of 150 are the strongest shower on the calendar.

For the Perseids specifically, the next maximum with a comparably dark sky is 2029, at 10% lit. That is a long wait, which is the whole argument for taking 2026 seriously.

Before the night

The full 2026 Perseid page carries the Moon across the whole activity window and the week around the peak. The front page computes the verdict for your own coordinates on the night, including the cloud forecast hour by hour, and tells you the next good night if that one is hopeless.