How to Use Non-Diatonic Chords to Great Effect in Your Songs

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

A non-diationic chord is one which does not normally belong to your key of choice. In the key of A major, the seven naturally-occuring chords are: A, Bm, C#m, D, E, F#m, and G#dim. Of those chords, there are countless ways of modifying them by adding tones (Aadd9, for example), using non-chord-tones (Esus4, for example), and so on. Of those seven chords, the ones you would find yourself using the most are probably I, IV and V (A, D and E in our example of A major)

Then there is a different category of chords called non-diatonic chords. These are ones which cannot be formed by building simple triads on the notes of the scale. Some of these chords have categories all their own. For example, secondary dominant chords are formed when you take a chord that is normally minor in a certain key and making it major, then following it with a chord four notes higher. Here’s an example:

A  F#m  B  E  A

The B chord is the secondary dominant. You would expect that chord to be Bm in our key of A major. By making it a major chord, then following it with a chord four notes higher (E), you’ve created a secondary dominant.

I want to focus, however, on the non-diatonic chords that don’t have roots that exist in the key. For example, the flat-3 chord. In A-major, the flat-3 chord is C. It can have a somewhat startling sound:

A  E  F#m  A  C  D  A

The great thing about the flat-3 is that it adds a certain moody colour to the sound of your progression. But here’s a way that it can have even better effect: Use it sparingly.

Keep in mind that chord progressions represent a type of groove that your song needs. For many songwriters creating chord progressions, once they’ve created one chord, they look for a new one, then another, then another, and eventually their progressions sound like a bit of a mess! The more chords you add to your progression, the greater the danger of killing the sense of groove in your songs.

So especially if you plan to use a non-diatonic chord like a flat-3, set up a groove using two basic chords, maybe a third chord for variety, from your chosen key, go back and forth between those chords to establish the groove, and then throw in the flat-3. You’ll love the result:

A  Bm  A  Bm  A  Bm  A  D  A  Bm  A  Bm  C  D  A

That C chord jumps out and demands attention when it happens, but it does so in a really great way: Every groove needs a bit of shaking up just for variety, and that’s the power of the non-diatonic chord. So just throwing non-diatonic chords everywere may be a little like making a soup where you throw in every spice in your rack: none of them are being used to good effect.

So set up a chord groove, and then let that non-diatonic chord pleasantly disrupt proceedings!

If you want to read about how to solve your songwriting woes, get Gary’s suite of 5 songwriting e-books at a “bundle discount” price. Click here to learn more..

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


  1. Is there something special about the flat-3, i.e. does it make more sense than other choices? Or did you just pick it on random as an example?

    If certain non-diatonic chords make more sense than others, what is the theoretical explanation for this?
    Thanks for great lessons btw!

    • Hi Andreas:

      That’s a really great question! Yes, I think there are reasons that certain non-diatonic chords appear more often than others. The main theoretical reasons are hard to relate in one reply, but I’ll give it a try. It involves how far from the home key the non-diationic chord in question is. Let’s take the example of C major. The chords that are most used, other than the ones that occur naturally in the key, are the flat-7 (Bb), flat-3 (Eb) and the flat-6 Ab. If you think of this in terms of key signatures, Bb has 2 flats, Eb has 3, and A flat has 4 flats. Beyond that, the non-diatonic chords get a bit difficult to use because they are theoretically far from the home key. So flat-2 is not as common, and flat 5 is even rarer.

      Also, chords that move by semitones are not very common in music, with the possible exception of the iii-chord moving to IV. So once you’ve used flat-2 or flat-5, it’s hard to know where to go from there.

      There’s a fair bit more one can say about this really interesting point you bring up, but hopefully that at least mostly answers your question.

      Thanks very much for writing,

      • So the three “flat-chords” you speak of can be seen as the tonic chords of the three scales that are harmonically closest to the current scale (that is, closest in the counter-clockwise direction in the circle of fifths). Makes a lot of sense, thanks!

        One question though, if I may: What about clockwise? Wouldn’t the above mean that sharp-5, sharp-2 and sharp-6 are just as valid as non-diatonic chords as flat-7, flat-3 and flat-6, respectively?


        • The enharmonic equivalents that you mention don’t really exist, since the sharp-5 would theoretically be a key with 8 sharps (i.e., the key of G major plus one semitone). The same for the other keys you listed.

      • I think it’s worth noting that the flat III, VI, and VII can be thought of as the 3 major chords that occur if the key was minor instead of major (borrowed chords). Same with making the IV chord minor. They all sound amazing.

        Also, about chords moving in semitones… in jazz when you use tritone substitutions after making minor chords dominant, you can literally descend the same dominant chord 1 fret several times. So a III-VI-II-V-I in the key of C could become E9-Eb9-D9-Db9-C.

        I don’t know much about chord substitutions since I’m learning myself but thought I’d mention it.

    • For my money the most common (for good reason) non-diatonic chords are (more or less in order): 1) the aformentioned Secondary Dominants; 2) the III chord, but major; 3) the IV chord, but minor, and the bVII chord (major, dominant 7, or sus).

      On another note, I would argue maybe a just a bit the veracity of chords that move by half-step. Just referring to root chords (inversions make for a great number of chord changes that move by half-step, at least in the bass), you can always precede a chord above or below by a half-step with the same quality of chord. Similarly, you can always throw a diminished 7 chord in between two diatonic chords, e.g., A – A#dim7 – Bm – E7. For bridging the gap between IV and V, I like the half-diminished 7 chord, or for a blues/blues-rock a regular Dominant 7 chord. This technique tends to sound Jazzy when used a lot, but can sound really cool used in modern pop songs when done sparingly. Bands like Muse make use of a lot chord progressions whose theory is rooted in classical and jazz music.

    • Most songs use chords that point more-or-less clearly to one tonal centre, so yes, most songs use diatonic chord formulas. Having said that, it’s also true that many songs will also include chords that are either borrowed from other keys, or that use chromatic alterations.


  2. Hi there!
    I am studying music technology at the moment in college and our current module is the study of Non-Diagnostic Chords. What you wrote above seems to make sence to me but I was wondering if you could explain it more. If you are busy no worries but any more info on the subject would be greatly appreciated.
    Kind regards,

Leave a Reply

Your email address will not be published.

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