Inside a Mahadasha: Antardasha, Pratyantardasha, and why the dates slide

The Mahadasha article stopped at the Antardasha: Saturn's sub-period inside Saturn's own Mahadasha runs 19 x 19/120 years. This one takes that arithmetic a level further down, then asks the question that decides whether any of it is usable: how far do the dates move if your birth time is slightly wrong?

One rule, applied twice

To cut a period into sub-periods, give each of the nine lords a share of its length equal to that lord's own Vimshottari years out of 120, and hand the shares out in the fixed cycle order, starting with the lord of the period being cut. Brihat Parashara Hora Shastra gives both halves in Chapter 51, as the citations in this engine's code and its research notes give it: Sloka 1 for the multiplication, Sloka 2 for the order - "the first Antar Dasha belongs to the Lord of the Dasha".

Cut a Mahadasha that way and you get nine Antardashas. Cut an Antardasha the same way and you get nine Pratyantardashas. In this engine that is one function called at both levels, and each Pratyantardasha is cut from its parent's real dates rather than recomputed from the chart. The code cites no separate verse for the third level: it is the second level's rule, applied again.

So inside a full-length Mahadasha, a Pratyantardasha lasts three year-counts multiplied together and divided by 120 twice. Saturn inside Saturn inside Saturn: 19 x 19 x 19 / 14,400 = 0.476 years, about 174 days.

What the verses actually say

Read in the translation this site works from, Chapter 51 is more explicit than the engine's citation suggests. Sloka 1 gives the rule for both levels in one breath: for an Antardasha, multiply the two lords' dasha years and divide by the total of all the lords' years; for a Pratyantardasha, multiply the length of the Antardasha by the other lord's years and divide by the same total. Sloka 2 gives the order - the first sub-period belongs to the lord of the period, the other eight follow in the dasha order - and ends with one short sentence: "The same applies to Pratyantar Dasha."

Chapter 61, which describes the effects of Pratyantardashas, opens by stating the rule a third time, now with the number written out: multiply the Antardasha by each lord's years and divide by 120. So the engine's "one function at both levels" is not an engineering shortcut that happens to agree with the text. It is the text's own description: one rule, applied again.

It is worth noticing what Chapter 51 does not do. The same chapter gives quite different rules for the sign-based dashas - their sub-periods are found by dividing the period by twelve, not in proportion to year-counts, and their order depends on whether the sign is odd or even, movable, fixed or dual. (This app's Chara Dasha does not use that order; it follows K.N. Rao's, starting from the sign after the period's own and ending with it.) The proportional nine-way split is specific to the nakshatra dashas like Vimshottari. Anything you read about Antardashas in a Chara Dasha follows a different arithmetic entirely.

If you use more than one dasha system, keep their sub-periods apart in your head. The Chara Dasha calculator and the alternate dashas calculator compute other systems, each with its own rules; the twelve-way sign split and the nine-way proportional split do not line up, and a date that is a boundary in one is usually not a boundary in the other. That is not a contradiction between them. They are different clocks, and the text presents them as such.

The nine shares, as fractions of any parent

Because the rule is a pure proportion, every lord takes the same share of every parent, whatever the parent is. It helps to have those shares in front of you once:

Two consequences follow. First, Venus, Saturn, Rahu and Mercury between them take more than three fifths of every period, at every level; the Sun, Ketu and Mars take about a sixth together. So most of the time, at any level, one of the four long lords is running. Second, the rule is symmetrical. The Saturn sub-period inside a Venus Mahadasha is 20 x 19 / 120 years, and the Venus sub-period inside a Saturn Mahadasha is 19 x 20 / 120 years - the same three years and two months. A pairing of two lords lasts the same length whichever of them is the parent.

Why the period's own lord goes first

The verse the engine cites states this rule without arguing for it, and this article will not invent a reason. What the rule does to the structure is checkable. Every period opens with its own lord: a full-length Saturn Mahadasha opens with a Saturn Antardasha of three years and three days. And every Mahadasha closes with the lord that sits just before it in the cycle, which after the first Mahadasha is the lord of the one before it. On the calculator's sample chart, Mercury's Mahadasha (2020 to 2037) ends in a Saturn Antardasha, and the Mahadasha before it was Saturn's.

The actual lengths

Of the 729 Pratyantardashas in a full cycle, 235 last under 30 days. That is the case for the third level in one number. "Mercury Mahadasha" covers seventeen years. "Mercury-Sun" narrows it to about ten months. "Mercury-Sun-Rahu" narrows it to about six and a half weeks.

Levels four and five, and why the app stops at three

BPHS does not stop at the third level either. Chapter 62 cuts each Pratyantardasha again into Sukshma dashas, and Chapter 63 cuts each Sukshma into Prana dashas, by the identical rule: multiply by each lord's years, divide by 120. Each chapter then lists an effect for every combination, just as Chapter 61 does for the third level.

Set those lengths next to the engine's own accuracy limit, described further down: an error of about 0.1 degree in the Moon slides every boundary by roughly 2.7 days per year of the opening lord. A Prana dasha lasting twenty minutes, or even five days, is smaller than that uncertainty. The arithmetic would happily print it; it would mean nothing. That is why this app computes three levels and no more, and why it asks you to read even the third as "which stretch, give or take weeks".

The stack, on one real chart

The calculator's own sample is a birth on 14 August 1994 with no time given, read at noon India time. On 30 September 2026 it returns Mercury Mahadasha (April 2020 to April 2037) and, inside it, the Sun Antardasha from 17 July 2026 to 23 May 2027: 17 x 6/120 = 0.85 years, 310.5 days. The calculator stops there. Here is the next level, worked by hand, matching what the engine computes:

Signed in, the Intelligence tab of the Kundali page shows all three running levels and how far through each one you are. A level the engine cannot place comes back empty, never as a guessed lord.

A second chart: a whole Mahadasha, cut nine ways

The Mahadasha essay works through a birth on 15 August 1990 at 10:30 IST in New Delhi, with the Moon in Rohini. That chart entered a 16-year Jupiter Mahadasha on 5 December 2018. Here is that whole Mahadasha, as the engine divides it:

Two things are visible here that the rule implies but a reader rarely notices. First, the sub-periods are very uneven: the Sun's is under ten months and Venus's is nearly three years, inside the same Mahadasha. Second, the sub-period that ends the Mahadasha - Rahu - is the lord that comes just before Jupiter in the cycle, and Rahu was the Mahadasha this chart had just come out of. A Mahadasha always closes on a sub-period of the lord before it, which gives the handover between Mahadashas a kind of echo.

How the classical text reads a sub-period

Dates are the half of the system that can be computed exactly. The other half is meaning, and here BPHS spends nine whole chapters - 52 to 60, one for each Mahadasha lord - describing the effect of every Antardasha inside it. The descriptions are long and vivid, and it is easy to quote them as if they were fixed predictions. They are not written that way. Almost every one is conditional, and the conditions follow one pattern.

Take Chapter 52, the Sun's Mahadasha. The Sun's own sub-period is good, it says, if the Sun is exalted, in its own sign, in the 11th, in a kendra or in a trikona; adverse if debilitated or badly placed; middling otherwise. The Moon's sub-period inside the Sun's period is described with domestic gains and property if the Moon is in a kendra or trikona, and with friction and losses if the Moon is waning, joined with malefics, or in the 6th, 8th or 12th. Then a second test is added: benefics in the 1st, the 9th or a kendra "from the Lord of the Dasha" bring good results, and the Moon in the 6th, 8th or 12th from the Lord of the Dasha brings poor ones.

That phrase - "from the Lord of the Dasha" - recurs dozens of times across Chapters 52 to 60. It is the central idea of sub-period reading in the text. A sub-period lord is judged twice: once on its own condition in the chart, and once by where it stands counted from the Mahadasha lord. Placed in a kendra or trikona from the Mahadasha lord, it is read as working with it; placed 6th, 8th or 12th from it, as working against it. Two people in the same "Jupiter-Saturn" period can therefore be in very different situations, depending on the distance between Jupiter and Saturn in their charts.

Chapter 61 states the principle outright at the third level. After describing the Sun-Sun Pratyantardasha with arguments and losses, it says those "inauspicious effects will not be produced" if the Sun is in a trikona, rules a good house, or sits in a benefic division - and adds that "all other Pratyantar effects should be judged in this manner". The lists are the general case; the chart decides.

To apply the "from the Lord of the Dasha" test yourself, you need the house of each planet, which the birth chart calculator gives. Count from the Mahadasha lord's house to the sub-lord's house, inclusively. A count of 1, 4, 7 or 10 is a kendra; 5 or 9 a trikona; 6, 8 or 12 the difficult positions the text warns about. The house pages for the 6th, 8th and 12th explain why those three are treated as they are.

The engine's own timing reading does not apply this test. As the Mahadasha essay explains, the convergence reading classifies a lord by its nature alone. The dates it gives are exact arithmetic; the tone it gives is the general case. The pages for each lord - for instance Jupiter Mahadasha, Saturn Mahadasha and Venus Mahadasha - describe what the texts say about the sub-periods inside each.

The test applied to the 1990 chart

In the 1990 chart, the ascendant is Virgo and the houses are counted whole-sign, so Cancer is the 11th house. Jupiter, the Mahadasha lord, sits there in Cancer - its sign of exaltation - together with the Sun, Venus and Ketu. Counting from Jupiter's sign to each sub-period lord's sign gives this:

This is not a reading of the chart. It is only one of the tests BPHS stacks on top of each other - the sub-lord's own dignity, its house from the ascendant, its lordships and the aspects on it all count too - and a 6th-from position can be outweighed by any of them. It shows how the test is actually performed, and why "Jupiter-Saturn" cannot have one fixed meaning across charts. In a chart with Saturn in a trikona from Jupiter, the same named period is read the other way.

A modern practitioner says the same, and adds transit

The same two-lord test survives in modern case-study writing. Advanced Predictive Astrology, a 2000 collection of worked charts that uses "Bhukti" for what this page calls the Antardasha, says that when the dasha and sub-period lords sit 6/8 or 2/12 from each other, unfavourable results are indicated, and that the same dasa-bhukti principle can be applied to the divisional charts. It lists what a running pair should be judged on - significations, house ownership, association, aspect, dignity, combustion, retrogression, and "the relation between the Dasa planet and the Bhukti Dasa planet".

It also makes a point Parashara makes more briefly: dasha alone is not enough, and the transits at the time have to be read on top. Its image for this is a clock, with dasha as the hour hand and transit as the minute hand. The book makes accuracy claims for that method that this page does not repeat, because nothing here can check them. The useful part is the structure: a sub-period says which planet's agenda is running, and the current transits say what is being pressed on at the moment. The timing feature sets the dasha next to the app's other timing layers for the same reason.

What each level is good for

It helps to be plain about what scale of question each level can answer, because most confusion about dashas comes from asking a level a question it is too coarse, or too fine, to answer.

• The Mahadasha is a background. Six to twenty years is the scale of a career phase or a stage of life. It says which planet's agenda the period is organised around, and nothing about a given month.
• The Antardasha is a theme. Median length a little over a year; it is the level at which the classical texts describe concrete matters - property, study, partnerships, disputes - and the level at which the "from the Lord of the Dasha" test applies.
• The Pratyantardasha is a window. Median about seven weeks. It is useful for noticing when a theme comes forward, and for checking a timing claim against a diary. It is not precise enough to pick a day, for the reasons in the note above.
• Below that, the arithmetic continues but the chart data does not support it. The app does not compute it.

Chapter 61 itself reflects that difference in scale. Where Chapters 52 to 60 spend a paragraph or more on each Antardasha, Chapter 61 gives each Pratyantardasha a single line - a handful of words for each of the nine inside each of the 81 sub-periods. The Sun-Sun line, for instance, is a short list of frictions and losses; the Moon-Moon line a short list of gains in land and property. The brevity is the point: the third level is a modulation of the second, not a story of its own, and the text immediately tells the reader to adjust each line by placement.

Working the third level by hand

The calculator prints two levels. The third takes about five minutes with a phone calculator, and doing it once is the best way to see that nothing in it is hidden.

1. 1.From the calculator, write down the running Antardasha: its lord, its start date and its end date.
2. 2.Count the days between the start and the end. That is the length you are dividing.
3. 3.Write the nine lords in cycle order - Ketu, Venus, Sun, Moon, Mars, Rahu, Jupiter, Saturn, Mercury - and rotate the list so it starts with the Antardasha's own lord.
4. 4.For each lord in that rotated list, multiply the length in days by the lord's years (7, 20, 6, 10, 7, 18, 16, 19 or 17) and divide by 120.
5. 5.Add the results one after another to the start date. Each running total is the date one Pratyantardasha hands over to the next; the ninth should land on the Antardasha's end date, give or take rounding.

The table of the 1994 sample above was made exactly that way: 310.46 days, multiplied by each share. The check in step 5 matters. If your ninth date does not land on the end date, a lord was skipped or the list was rotated to the wrong starting point - by far the commonest mistake, and the same one that BPHS's "the first Antar Dasha belongs to the Lord of the Dasha" exists to prevent.

What a wrong birth time does to the boundaries

The lengths above are exact arithmetic. The dates are not, because they all hang from the birth moment and one number: how far the Moon had travelled through its nakshatra at birth. That sets the balance of the first Mahadasha, and every later boundary is counted from where that balance runs out.

So as long as the Moon stays in the same nakshatra, an error does not stretch or shrink anything after the first Mahadasha. It slides all of it as one block. On the sample chart, a birth ten minutes after noon moves every later boundary, at all three levels, 42.9 days earlier, and the Pratyantardasha running on 30 September 2026 becomes Jupiter instead of Rahu. An hour after noon moves them 258 days, and the running Antardasha becomes Moon instead of Sun.

Measured on this engine across ten birth dates, each hour of birth time was worth roughly 14 to 17 days per year of the opening lord's period, depending on how fast the Moon was moving that day: between about three and a half and ten months per hour.

With no birth time at all it is worse, even on the dates where the ruling planet holds all day (18 of 395 dates sampled between 1970 and 2009). There the boundaries still spread across 88 to 98 percent of the opening lord's entire period, because the Moon spent the whole day crossing that one nakshatra. What an unknown birth time costs covers the dates where the planet itself changes.

The size of that choice is easy to underestimate. On the 1994 sample, the first Mahadasha is a partial Jupiter period, and the engine's Jupiter-Jupiter sub-period inside it lasts 325.8 days. A full-length Jupiter-Jupiter sub-period would last 779.2 days. The alternative convention - keep full-length sub-periods and start the clock part-way through them - would give a different sub-lord at birth and different dates for every sub-period until the first Mahadasha ends. Neither convention changes anything after that, since every later Mahadasha is full length. If you are a child or teenager, or reading a child's chart, this is the one place where two careful programs can disagree on the Antardasha without either being wrong about the arithmetic.

The same split applied to a nakshatra's 13 degrees 20 minutes instead of to years gives the sub-lord - an extension of the Chapter 51 rule, not a use the text itself makes.

Run yours

The Vimshottari Dasha calculator needs only a birth date, and a birth time sharpens every date it prints. It returns your nine Mahadashas and the nine Antardashas of the one running today. For the third level, take your running Antardasha's length and hand it out in the shares above, in the same cycle order, starting from the Antardasha's own lord.

If the ruling planet itself is uncertain because your Moon sits near the edge of a nakshatra, start one step earlier: the nakshatra calculator shows where the Moon is and how close to a boundary, and the Moon in Vedic astrology explains why the whole timing system hangs from that one planet.

Is "Bhukti" the same as Antardasha?

Yes. Bhukti and Antara are other names in common use for the second level. BPHS itself uses Antar Dasha, and Pratyantar Dasha for the third.

Why does the calculator not show Pratyantardashas?

The public calculator stops at two levels; the signed-in Kundali page shows all three running levels. The third level can be worked by hand from the second in a few minutes, with the steps above.

Why not go to the fourth and fifth levels, which BPHS describes?

Because they are shorter than the engine's own uncertainty. A Sukshma dasha can be hours long and a Prana dasha minutes; a 0.1-degree error in the Moon moves boundaries by days.

Do the sub-periods of the same lord always mean the same thing?

Not in BPHS. Chapters 52 to 61 judge a sub-period by the sub-lord's own condition and by its position from the Mahadasha lord, and Chapter 61 says the general effects are not produced when the placement contradicts them.

Does a longer sub-period mean a more important one?

No. Length comes only from the lord's year-count. A Venus sub-period is long and a Sun sub-period short in every Mahadasha, in every chart; what BPHS reads as significant is the sub-lord's condition and its position from the Mahadasha lord, which length says nothing about.

Is a sub-period of the Mahadasha lord itself special?

It is always the first one, by Chapter 51 Sloka 2, and it is the only sub-period in which one planet rules both levels. BPHS reads it, like the others, by that planet's dignity and house - Chapter 52 describes the Sun-Sun sub-period as good when the Sun is strong and adverse when it is debilitated.

Why do my sub-period dates differ from another app's by weeks?

A different birth time, ayanamsa or year length moves every boundary as one block, and a first-Mahadasha convention can change the sub-periods of childhood outright.