The Descending-by-Thirds Chord Progression

by Gary Ewer, from “The Essential Secrets of Songwriting” website:

There’s nothing random about a good chord progression. Starting on any chord, descending by 3rds will give you a great one to try.

Let’s use the key of A major as our example. Starting on the tonic chord (A), then allowing the progression to descend by 3rds, you’ll come up with a progression that works really well:

A  F#m  D  Bm… then finish with E  A

I think you’ll find that the progression isn’t quite as strong when done in reverse:

A  C#m  E  G#dim  Bm  D… finishing with E  A to get you out of the cycle.

The descending version of the progression feels more natural, and each successive chord feels instinctive and expected, in the best sense of those words. The ascending version feels random and forced. Here’s why:

Chords are classified by what we call their function. The I-chord (A) is a tonic chord, which feels like “home.” The V-chord (E) is a dominant chord: it likes to move to I. The IV and ii chords (D and Bm) are predominant chords: they like to move to dominant chords. Beyond the predominant chords are “the other ones”: iii-chords (C# minor) like to move to vi-chords (F#m) because of the root movement of a 4th. In general, vi-chords like to move to predominant chords. If you put all of what I just said in a chart, it would look like this:

iii (C#m)
vi (F#m)
PreDom: ii (Bm)   IV (D)
Dom: V (E) vii (G# dim)
Tonic: I (A)

The best way to read this chart is from top to bottom. In other words, start on the I-chord, then leap to any chord in that list, and move downward. You’ll always get a progression that works, and that is very strong. In the chart above, the chords nearer the bottom of the chart are more common and more used than chords nearer the top in most genres of music composition.

Pop music has a lot more leeway to do progressions outside of what this chart suggests, and in fact, in order to keep your music from being too predictable and boring, you’ll want to come up with progressions that don’t actually follow this chart all the time.

But back to the descending-by-thirds progression. It works better than the ascending-by-thirds progression because the ascending version is contrary to almost all of the natural tendancies of good chord progressions.

The descending progression (in Roman numerals) is: I  vi  IV  ii  V… all reflected nicely in the chart.

The ascending progression is: I  iii  V  vii  ii  IV… feeling a fair bit random because it doesn’t click very well with this chart.

Try using that chart above to rate the strength of your progressions. You’ll find that because the chart basically gives you strong progressions, those progressions will work better with choruses than with verses. And as I say, if you stick with this chart 100% of the time, you’ll wind up with very predictable, and perhaps boring, progressions. Balance predictable with innovative, and your songs should be fine.

If you’d like more information about chord progressions, including pages of progressions you can use right now in your own songsread about Gary Ewer’s songwriting e-books.

Posted in Chord Progressions, songwriting and tagged , , , .


  1. Really, ii – IV – vi – I, IV – vi – I – ii, and vi – I – ii – IV (the one the poster above mentioned) are all sound relatively nice to me. The first is basically the descending progression in reverse. And the first and the third progression are also, interestingly enough, the same chord progression, in relation to their starting chord (i – III – V – vii). Perhaps that has something to do with the palatability of both of them? The ascending progression is sort of meandering and pretty, and is somewhat disrupted by the presence of the diminished chords when the ascension is started from I, iii, V, and vii. Perhaps I am merely overly influenced by pop music, which seems to avoid diminished triads like the plague.

    Also, wouldn’t the descending progression be I – vi – IV – ii – vii, not I – vi – IV – ii – V?

    • Yes, the vii-chord would complete the descending-by-thrids pattern, but the point was that eventually there will be a need to break out of the thirds pattern in order to give a dominant chord that moves easily back to the tonic.

  2. Trying to teach myself music theory and am having trouble with a progression that seems to work but does not hold true to typically espoused progression theory. The progression is Am C Em G which is an ascending-by-thirds pattern yet doesn’t sound random, perhaps fragile but certainly not random. Could this be due to the fact that Am/Em and C/G are respective fifths and Am/C and Em/G are minor/major equivelents? In short, why the heck does this progression work?

    • Hi Garrett – The progression is, as you point out, “fragile”, for the reason that no one chord is strongly indicated as a tonic. Fragile progressions can be quite lovely, and while this one isn’t strong, there are reasons (and circumstances) where it will work well. The fifth relationship that you point out is a possible reason for it working, but what I think is more obvious is the relationship between adjacent roots of the chords. Since each chord is followed by one that has a root a 3rd higher, it strengthens the predictability of the progression. The descending version of that progression is stronger, as I mention in the post, but your version, while not as strong, can be quite nice.

      Thanks for writing,

      • also because of the shared harmonies of each of the chords…i.e AM has a C E….Cmaj has a C E…Cmaj has an E and G…Emin has an E and G..etc…your’re only ever changing one note..

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.