Moon phases and the lunar calendar: the cycle behind the dates

Moon phase · elongation from the Sun, 0=new, 0.5=full · synodic month about 29.5 days · 4 named phases about 1.8 days wide, the other 4 about 5.6 days wide

01Two different clocks, one word

A moon sign and a moon phase both get called "the Moon" in casual conversation, and they are answers to two different questions, measured two different ways. A moon sign asks where the Moon sits against the fixed background of the zodiac — a sidereal-style question, answered from the Moon's own longitude alone, on roughly a 28-day cycle to return to the same point against the stars. A moon phase asks something else entirely: not where the Moon is against the zodiac, but where it is relative to the Sun, as seen from Earth. That is a synodic question, and it runs on a slightly longer cycle, about 29.5 days from one new moon to the next, because the Sun's own position has crept forward during that month too and the Moon needs a little extra time to catch back up to the same angle from it.

That gap between 28 and 29.5 days is not a rounding difference to smooth over. It is the entire reason a phase calendar and a sign calendar drift apart from each other over time even though both start from the same Moon: one is timed against the stars, the other against the Sun, and the two only line back up exactly once in a while rather than every month.

Both figures are averages, worth stating as such rather than as fixed constants. The Moon's own orbit is an ellipse, not a circle, so its actual speed varies across a month — faster closer to Earth, slower farther away — and the real interval between two consecutive new moons can run a little short of 29.5 days or a little past it depending on where in that ellipse the Moon happens to be. 29.5 is the right number to reason with for almost any given month, the same way this product states the Ascendant's four-minutes-per-degree rate or the Moon's thirteen-degrees-a-day sign motion as working averages rather than constants that hold to the same precision every single time.

02What a phase actually measures

The number behind a moon phase is the Moon's elongation from the Sun — the angle between them, as seen from Earth — expressed as a fraction of a full circle running from 0 at the new moon to 0.5 at the full moon and back around to 1, which wraps to the same 0 as the next new moon. That fraction is not the same thing as how much of the Moon's disc looks lit, even though the two get treated as interchangeable in a lot of everyday phase widgets.

The illuminated fraction is a different calculation, and it has a real blind spot the elongation angle does not: because the amount of the Moon's face that catches sunlight depends only on how far around its orbit it has travelled, not on which direction it travelled to get there, a Moon three-quarters of the way to full from new and a Moon three-quarters of the way back down to new afterward can show an identical percentage lit while sitting on opposite sides of the same cycle — a waxing gibbous and a waning gibbous, indistinguishable by illumination alone. The elongation angle does not have that blind spot, because it runs the whole way around the circle rather than folding in half at the full moon: a waxing gibbous and a waning gibbous sit at two different elongation values even when they would report the same percentage lit, which is the whole reason the actual phase — not just how full the disc looks — is computed from the angle rather than from the brightness.

03Eight names, and how wide each one is

The everyday vocabulary of a phase calendar — new, waxing crescent, first quarter, waxing gibbous, full, waning gibbous, last quarter, waning crescent — names eight bands along that same 0-to-1 cycle, and the eight bands are not equal widths. Four of the names sit on an exact instant: new at 0, first quarter at a quarter of the way round, full at the halfway point, last quarter at three-quarters. Each of those four claims a narrow band centred on its exact instant, about three hundredths of the full cycle to either side — a little under two days out of a 29.5-day month — before the name changes to whichever crescent or gibbous phase comes next. The four names in between — the two crescents and the two gibbous phases — each cover a much wider stretch, close to a fifth of the whole cycle, something like five and a half days apiece, because there is no single exact instant inside a crescent or a gibbous phase for a band to centre on; the name simply applies for as long as the Moon is somewhere between one exact phase and the next.

Put in days rather than fractions: roughly two days carrying the name "full moon," and roughly five and a half days carrying the name "waxing gibbous" on the way there. That asymmetry is not an approximation chosen for convenience — it falls directly out of dividing one continuous cycle into four points that can be dated exactly and four stretches that, by definition, cannot.

04The calendar that never quite fits the year

Twelve of those 29.5-day cycles add up to about 354 days, roughly eleven days short of the 365.25-day cycle the Sun's own return to the same point takes. That gap is the entire reason a calendar built purely from lunar months drifts against the seasons rather than lining up with them year after year: twelve new moons come and go about eleven days earlier, in solar terms, each year that passes, so a date fixed to the eleventh new moon of a given year slides slowly backward through the solar calendar rather than landing on the same solar date twice. A calendar that wants to track both the Moon's phase and the Sun's season at once has to insert an extra correction somewhere to keep the two from drifting apart indefinitely; a calendar that tracks the Moon's phase alone simply lets the drift happen, eleven days a year, compounding.

05Finding the exact moment

Locating a real new moon, full moon or quarter on a specific date is not a lookup against a fixed table — it is found the same way any other exact astronomical moment in this engine is found, by sampling the Moon's elongation from the Sun day by day across a window, noticing where that value crosses one of the four target angles, and then narrowing in on the crossing by repeated bisection until it is pinned down to about a second. New and full moons are treated as the more significant pair of the four — weighted higher than the two quarters — because they are the two extremes of the cycle rather than its midpoints, the moments the Sun and Moon are either aligned or squarely opposed rather than at a right angle to each other.

That same scan, run across a requested window, is what actually powers a real calendar page rather than a static list: asked for a given month, it returns every ingress, retrograde station and lunation that falls inside it, computed fresh from the ephemeris rather than authored in advance, which is what makes it honest to point a reader toward as a page that really does compute the phases described here rather than merely listing them from memory. The same scan also assigns each event a significance value on the same 0-to-1 scale an aspect's strength uses, so the new and full moons at either end of the cycle are marked well above the two quarters that sit between them — a rarity-weighted ranking rather than a flat list, built on the same underlying idea as the strength score that ranks a chart's own aspects: something worth noticing stands out from something merely present.

06What it is not

A moon phase is not a moon sign, and the two should not be read as the same fact answered two different ways. Someone can be born under an exact full moon and have that Moon sitting in any of the twelve signs depending on where in the zodiac the full moon in question happened to fall — the phase says where the Moon sits relative to the Sun; the sign says where it sits relative to the fixed zodiac, and knowing one does not hand over the other.

An illuminated-percentage display, the kind a phone's weather widget often shows, is not wrong, but it is a narrower and blunter number than the phase itself: it can tell a full moon from a new one, but it genuinely cannot tell a waxing gibbous from a waning one on its own, for the reason worked through above. And a phase calendar is not an eclipse calendar. A new or full moon is a necessary condition for an eclipse but nowhere near a sufficient one — an eclipse additionally needs that new or full moon to fall close to one of the two points where the Moon's own orbit crosses the Sun's apparent path, which is a separate piece of geometry with its own separate article, and most new and full moons pass without one.

07Where you see it in your own chart

The calendar built from this same scan is computed fresh for whichever month is asked for, listing the lunations that fall inside it alongside the ingresses and retrograde stations happening at the same time, rather than a page of dates typed in once and left to go stale.

→ See this month's phases

08Related reading

The lunar nodes are the other half of the eclipse question this piece deliberately sets aside — the two points a phase alone cannot account for. The moon sign piece covers the sidereal half of the Moon's two clocks in full, for the placement a phase calendar does not touch. Cancer, the sign the Moon rules, is worth reading alongside both.

Updated 2026-05-13.