Kaal Sarp is a geometric condition with an unusually clean definition. Rahu and Ketu sit exactly opposite each other, so the line between them cuts the zodiac into two halves. The condition holds when all seven of the other grahas fall in the same half.
The calculator above tests it, names the form, and tells you how often it actually occurs. It also tests both halves, which sounds obvious and is where a lot of calculators quietly go wrong.
The definition, precisely
Rahu and Ketu are the two lunar nodes, the points where the Moon's orbit crosses the Sun's apparent path. They are always 180° apart, by definition rather than by coincidence, because they are opposite ends of one line.
That line divides the ecliptic into two arcs of 180° each. Call them the arc running forward from Rahu to Ketu, and the arc running forward from Ketu to Rahu.
Kaal Sarp holds when the Sun, Moon, Mars, Mercury, Jupiter, Venus and Saturn all lie in one of those two arcs, with none in the other.
Two things follow immediately. The nodes themselves are not counted, since they are the boundary. And both arcs count equally — a chart with everything crammed between Ketu and Rahu is exactly as much a Kaal Sarp chart as one with everything between Rahu and Ketu.
The mistake that halves the rate
That last point is the one to check in any calculator you use.
Several implementations test only whether every graha lies within 180° measured forward from Rahu. That catches charts hemmed in on one side and misses every chart hemmed in on the other, and since the two cases are about equally likely, such a calculator finds roughly half the true number of cases.
The failure is invisible from the outside. It never produces a wrong answer for a chart it does catch; it simply reports "not present" for a chart that is present. Nobody notices a false negative on a condition they were hoping not to have.
This page reports which arc the grahas are hemmed into: Rahu forward to Ketu, or Ketu forward to Rahu. If a calculator does not tell you which side it found, it may not be looking at both.
How common is it?
Uncommon, and this is the interesting part.
Measured over four thousand real charts spread across sixty-five years, the full condition holds for about 7.5%. A near miss, with six of the seven inside and exactly one stray, occurs for about 26%.
That makes Kaal Sarp genuinely different from most of the classical doshas. Mangal dosha, on the common three-reference test, catches nearly ninety per cent of people. Sade Sati is running for about a quarter of everyone at any moment and reaches everybody eventually. Kaal Sarp catches about one person in thirteen, so a positive result actually narrows things down.
Whether it means anything is a separate question this page takes no position on. But a test that selects one in thirteen is at least carrying information, which is more than can be said for several of its neighbours.
Why it is rare, and why it is not rarer
The frequency is worth understanding, because it explains the shape of the result.
If the seven grahas were scattered independently around the zodiac, the chance of all seven landing in a specified half would be one in 2⁷, and doubling for either half gives about 1.6%. The measured rate is nearly five times that.
The reason is that the grahas are not independent. Mercury never appears more than about 28° from the Sun, and Venus never more than about 47°. So the Sun, Mercury and Venus travel as a loose group, and a chart that puts the Sun on one side of the nodal axis has already put two others there most of the time. That clustering makes a one-sided arrangement far more likely than random scattering would suggest.
The same clustering is why the near-miss rate is so high. Twenty-six per cent of charts have exactly one graha out of position, and the stray is very often the Moon, which is the fastest body and the least tied to the rest.
Checking the arithmetic on the rarity
The clustering argument is easy to verify roughly, without any simulation.
Start with the Sun. Wherever it is, Mercury is within 28° of it and Venus within 47°. So those three occupy a stretch of sky at most about 95° wide, which is barely more than a quarter of the zodiac and comfortably fits inside a 180° arc almost however it is placed.
That reduces the problem. Instead of seven independent bodies, you effectively have the Sun-Mercury-Venus group as one loose object, plus the Moon, Mars, Jupiter and Saturn moving freely. Five things to place rather than seven, and five things landing in a specified half is one in 32, doubled to one in 16 for either half, which is about 6%.
The measured 7.5% sits just above that, and the remaining gap comes from the fact that the outer planets move slowly and so are themselves somewhat clustered over any short span of years.
None of that is a claim about meaning. It is a check that the number the tool reports is the number the geometry implies, which is the sort of thing worth confirming before quoting a statistic at anyone.
The twelve named forms
When the condition holds, the form is named for the house Rahu occupies, counted from the lagna: Anant, Kulik, Vasuki, Shankhpal, Padma, Mahapadma, Takshak, Karkotak, Shankhachud, Ghatak, Vishdhar and Sheshnag for houses one through twelve.
Spellings vary between sources and the order occasionally does too, which is worth knowing if you are comparing two readings.
There is a structural point hiding in the naming. The condition itself depends only on planetary longitudes, so it needs nothing but a birth date and a rough time. The name depends on the house Rahu occupies, which depends on the lagna, which needs the birth time to the minute. So a chart can be confidently Kaal Sarp and only provisionally Takshak, and those two facts have different reliability. The calculator asks for a birth time for the naming, and would give you the yes-or-no answer without one.
Mean node, true node, and the charts on the boundary
One technical choice moves results here more than anywhere else on this site.
Rahu and Ketu can be computed as the mean node, a smooth average, or the true node, which oscillates around the mean by up to about 1.5°. This page uses the mean node, as most Indian practice does.
For Kaal Sarp specifically, that choice is decisive for a particular class of chart: one with a graha sitting close to the axis. A graha within 1.5° of Rahu or Ketu can be inside the arc under one convention and outside it under the other, flipping the whole result. Two sites can therefore disagree about whether you have Kaal Sarp without either having made an error.
If your chart has a graha near the axis, the answer is not robust, and that is worth knowing before anyone builds anything on it.
Why another site gives a different answer
Only one arc tested, as described above. The commonest real bug, and it produces false negatives only.
Mean against true node, worth up to 1.5° and decisive for charts with a graha near the axis.
Whether a graha exactly conjunct Rahu or Ketu counts as inside or outside. A judgement call that different implementations resolve differently.
Partial versus full. Some sources treat six-of-seven as a partial Kaal Sarp and report it as present; this page reports it separately, because 26% and 7.5% are very different findings.
The ayanamsa, which matters here less than usual. The condition compares grahas against Rahu, and the ayanamsa shifts everything equally, so it cancels almost entirely. It affects only the house Rahu occupies, and therefore the name rather than the finding.
That last point is worth dwelling on, because it is unusual. Most tests on this site move with the ayanamsa. This one is nearly immune to it, since it is built from differences between longitudes rather than from absolute positions. It is the same reason tithi is immune where nakshatra is not.
Reading a result you have been given
Four checks, in order, on any Kaal Sarp reading.
Which arc did it find? If the report does not say, it may only have tested one. Ask, or test elsewhere.
Is any graha within a degree or two of Rahu or Ketu? If so the result is fragile: the mean-node and true-node conventions will disagree about it, and so will two otherwise correct calculators.
Is it full or partial? Six of seven is a quarter of all charts and is a substantially weaker claim than seven of seven. Sources that report both under one heading are conflating a one-in-thirteen finding with a one-in-four one.
Was a birth time used, and does the claim depend on it? The yes-or-no answer does not need one. The named form does. If a reading leans on the name, it inherits every bit of your birth-time uncertainty; if it leans only on the condition, it does not.
The two nodes are not planets
One thing worth stating plainly, because a good deal of writing about Kaal Sarp obscures it.
Rahu and Ketu are not bodies. Nothing is there. They are the two points where the Moon's orbital plane crosses the plane of the Earth's orbit, and they matter astronomically because eclipses can only happen when the Sun and Moon are near them. That is where the mythology of a serpent swallowing the Sun comes from, and it is a genuinely sharp piece of observational reasoning: ancient astronomers identified the two invisible points on the sky that govern when eclipses occur, and named them.
They move backwards along the ecliptic, completing a circuit in about 18.6 years, which is why they are always marked retrograde. That regression is real motion of the orbital plane rather than an appearance, unlike planetary retrogradation.
So the Kaal Sarp condition is, stripped of language, a statement about where the seven visible bodies sit relative to the eclipse line. Whether that arrangement carries significance is not something this page can settle. What it can say is that the line itself is a real feature of the sky with a precise definition, which is more than can be said for some of the reference points other tests are built on.
What this page will not do
It will not tell you what a Kaal Sarp chart means for your life, and it will not sell you a puja to correct it.
What it computes is the angular position of each graha measured forward from Rahu, whether all seven fall within 180° of each other on one side, which side that is, and which house Rahu occupies. Every one of those is checkable by hand from an ephemeris and a protractor.
The one substantive thing worth carrying away is the frequency, and it cuts in an unfamiliar direction here. On this site's other dosha pages the base rate deflates the finding, because the tests catch most people. Here it does not. About one chart in thirteen qualifies, and if yours is one of them you are in a genuinely small group. What follows from that is not a question arithmetic can answer.