As far as i know, almost all songs end in the tonic chord, because this is the most "stable" condition. On the other hand, a dominant chord at the end of a progression (or song) sounds "incomplete".

However, i have found several songs that end with a V chord, and it sounds fine. (A Chinese folk song titled Mo Li Hua is one notable example). In my observation on these songs, i found a common thread: the last V chord is preceded by a II maj chord (i.e the secondary dominant V/V chord). Is this really a common characteristic of this group of songs (or is it a coincidence)... and what else would make "ending at a V chord" more "stable"?

  • As a follow up question, does this phenomenon apply to the natural minor scale as well? ie. Ending at v minor chord.
    – mey
    Jan 30, 2015 at 15:58
  • Another example of this is Shi Shang Zhi You Ma Ma Hao 世上只有妈妈好. Notice that both songs are in the pantatonic scale, so you can think of the V as the tonic from the beginning.
    – GZ-
    Jun 17, 2023 at 14:07

7 Answers 7


The making the ii chord major turns it into a secondary dominant of V (V/V) which makes V a temporary tonic. If you are ending on V in this case you would technically end on a tonic although it is not the tonic present though the rest of the piece.

I wouldn't go as far as calling ending on V unstable as it is a very typical cadence that can be seen in many pieces. The effect I would describe instead is that ending on V (or a half cadence) makes a piece seem like it's going to continue or there's more to the piece then there is. This isn't necessarily a bad thing and can even be used to create a type of "musical cliffhanger". I know I've used it that way in my own songwriting especially if I wanted the end of a song to be kind of ambiguous.

It should be noted however that the chord chosen to end a piece is only one of the things that constitutes the ending of a piece and how it overall feels. I guarantee that the same ending progression (cadence) can feel complete (stable) or incomplete (unstable) depending on how it is approached.


Preceding the V chord with a II chord is actually a perfect way to make it more stable. When using a secondary dominant chord, the song's key is momentarily substituted by the key of the fifth, which means that the II chord functions as a dominant chord for a moment. Hence the V/V notation, and the "secondary dominant" name. A logical consequence is that the following V chord sounds much more stable than it usually would (when preceded by a IV, for instance). In many cases, the V chord resolves to I, which results in two consecutive chords relieving tension.

That last part is not necessary though. It's perfectly alright to end a song with II followed by V, which is relatively stable.


Ending a tune on the dominant chord is a common technique for performers to keep the suspense and interest going in a given set. Usually that dominant chord will have some relationship to the key of the succeeding tune. (You are in C, end on G, then take off on a tune in G or D or G minor, etc.) It can be over-used, of course.

I have also used this device on the final song of the show, making for a neurologically disturbing conclusion to the performance that can reverberate for the audience "until next time."

That is different than when a series of changes at the end of a piece scrambles or changes the key feel so that the ostensible dominant chord really functions as a particularly thrilling concluding chord that somehow sounds right.

  • Thanks for the tip @memphisslim. Now I understand why this series of ending tune that i copied from a concert (which is E G A.. G C. DCDE.., ending in Em for an A minor song) sounds fine despite the incomplete cadence.
    – mey
    Jan 30, 2015 at 17:48
  • 1
    The E G A sequence introduces some modal possibilities that others here are more competent to discuss. I know that sequence of chords was popular during British Invasion days. E.g., the Yardbirds' For Your Love, which set the stage for others: I'm Not Your Stepping Stone, recorded by a bunch of people, and others. They did tend to reach for the dominant B7, but sometimes surprised by reaching for a C. We did not worry about what key we were in, but liked these songs insofar as they got us away from the I IV V sequences that characterize most Blues/Rock and Roll tunes we were playing. Jan 30, 2015 at 18:39
  • Sorry for using a confusing notation... when i wrote "E G A. .." etc, i was referring to the sequence of notes rather than chords. As far as i can remember, the last 3 chords for this sequence were vi, ii and iii.
    – mey
    Jan 30, 2015 at 18:44
  • 1
    Well, we are all communicating through a pin hole, and you are doing fine, Mey. Jan 30, 2015 at 18:46
  • Whoops, i made another mistake. Those chords were in reference to a major key, so in natural minor, they would have been i, iv and v. I have a strong tendency to think in major keys and must now get used to thinking in minor keys.
    – mey
    Jan 30, 2015 at 18:52

I don't think it's particularly uncommon for popular music to have both the verse and refrain end on a dominant chord, though in many cases the last verse or refrain will be followed by a (possibly instrumental) coda which either resolves to the tonic or fades to nothing.

When that pattern is used (each verse or chorus ends with a dominant that leads into the major or minor tonic at the start of the next verse or chorus), I don't think there's any need to have the tonic preceded by a secondary dominant.

Hotel California (The Eagles), for example, has a verse i V VII IV VI III iv V, and a refrain VI III V i VI III iv V. The song never really "lets up"; the V i cadence within the refrain isn't really a "resolution"; the main dominant-tonic resolutions are between the end of each verse and the start of the next. The piece ends by repeating the chords from verse and fading out.

Another example of a piece which ends the verse and bridge sections on a dominant is "On My Own" [Les Misérables]; those sections end on the dominant, which generally resolves to the tonic of the next verse/bridge except that the last verse jumps to a new key, and is followed by a coda which is similar to the start of the verse, but resolves on the tonic.

  • Thanks @supercat, so probably fading out is another useful strategy after finishing a song with a dominant chord. Worth trying for me :)
    – mey
    Jan 31, 2015 at 0:43
  • 2
    @mey: I think fading out is a "cheat", personally; I prefer to simply end with a tonic chord where the downbeat of the next verse would be, but fading out does avoid the need to resolve.
    – supercat
    Jan 31, 2015 at 2:37

In addition to Lee's answer, a Io (just for you Lee!) will sound good preceding the V, which could then end the song. It's not far off being a V/V, as in C, for example, V/V will be spelled D,F#,A,C - with a dominant 7th bit attached - whereas Io is C,Eb,Gb,Bbb (A). Which could almost be D7b9, another sort of V/V.

  • Interesting. . Probably the reason (why C dim 7 can replace D7b9) is that C dim 7 is essentially an inversion of D7b9 without the root?
    – mey
    Jan 30, 2015 at 15:53

A song can end IN the tonic key but ON the dominant chord. Or it can modulate to the dominant key and forget to come home! (This is standard practice of course at the end of the first SECTION of a song. A journey out, followed by a journey home.) In Western music, we use a lot of secondary dominants. A favourite trick is to start a song in C major, and stay close to that tonality for the first section. Then, after the double bar, jump straight into an E chord, and follow the circle of 5ths home - E7, A7, D7, G7, C. Each chord the dominant of the one following it. Don't read anything special into V being preceded by II(major)7. It's common, whether the sequence gets all the way home to I or not! You don't end on V because you want a feeling of stability! Quite the opposite.


Bach often harmonized chorale melodies that are in the Phrygian mode in the relative minor key, ending on the dominant. An example is the passion chorale (usually sung in English with the words "O Sacred Head Now Wounded"; also sung by Simon and Garfunkel). The melody is in E Phrygian, but the harmonization may be analyzed as A minor ending on the dominant. (Another option is to harmonize in C major with the melody ending on the third of the tonic chord. Bach set this chorale probably dozens of times.)

Similarly, a chorale in the Mixolydian mode will generally sound as though it is ending on the dominant chord of the relative major key. An example is Komm Gott Schöpfer, heiliger Geist.

In such contexts, the next-to-last chord is not likely to be the secondary dominant of the last chord.

  • 1
    Regarding modes and their varying interpretations, I feel obligated to drop this Adam Neely clip about the tonal center of Sweet Home Alabama. You know, for posterity and all :)
    – user45266
    Jan 14, 2020 at 1:20
  • @user45266 thanks for that. Very clear and entertaining too.
    – phoog
    Jan 14, 2020 at 1:24
  • 1
    @user45266: I appreciate the link too, and the video is informative. The main issue is that people start with the false assumption that modern pop harmony is tonal in the same sense that Mozart is tonal. Or that it's necessarily "in a key" in the same way the Mozart is in a key. Modern pop harmony just works differently, and Tagg isn't the only person who has been analyzing pop music in better ways. Unfortunately, most intro harmony books seem to parrot nonsense out of century-old theory that doesn't apply to pop music (and was never intended to model it).
    – Athanasius
    Jan 14, 2020 at 18:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.