For example, we talk about Beethoven's Symphony No. 5 in C minor, yet the movements are in a variety of keys:

  1. Allegro con brio (C minor)
  2. Andante con moto (A♭ major)
  3. Scherzo: Allegro (C minor)
  4. Allegro (C major)

The key can also change within a movement. For example, the first movement begins in C minor with the iconic "da-da-da-daaas", but Wikipedia states:

[the] second theme is in E♭ major, the relative major [to C minor], and it is more lyrical...

So now we have the additional task of determining the overall key of each movement.

Factors for determining overall key

Here are some factors that I think could determine (or at least influence) the overall key of the work:

  • The (first?) key of the opening movement
  • The (last?) key of the final movement*
  • The key used most often within the piece
  • The key used for the most iconic section of the piece
  • The key that any recurring theme within the piece is written in

I could construct a similar list of criteria for determining the overall key of each movement, but the question is: what is the deciding factor?

* Since Beethoven's 5th doesn't end in C minor, this would appear to go against the idea that the overall key is the final key. However, Wikipedia states:

The music [of the 4th movement] resounds in C major, an unusual choice by the composer as a symphony that begins in C minor is expected to finish in that key.

So if Beethoven's 5th is an exception then perhaps the final key is important in determining the overall key, at least most of the time?

Note: I have used Beethoven's 5th Symphony as an example, but people are welcome to refer to other pieces in their answers/comments. I am really interested in the general case, and links to evidence are appreciated.


Heinrich Schenker's notion of the "auxiliary cadence" (Hilfskadenz) starts to answer this very question: it tries to explain how a movement can begin away from tonic.

I don't have firm data for this, but my experience as a musician (and conversations with other musicians) tells me that the general belief is that a composition's final key is its overall key.

But interestingly, this only tends to be true for individual movements. When we get to multi-movement works, opinion seems to be focused on the beginning. (Again, I don't have data to prove this, this is just my experience.)

Originally, these multi-movement works were always bookended by the same key. The four movements of Haydn's "Surprise" Symphony, for instance, are in G, C, G, and G. This key---G major in this case---was viewed as the overall key.

Occasionally, composers would get a little innovative, often by giving a "slow introduction" to the first movement. Haydn's Symphony 104, for instance, begins in D minor. But the four movements still have a pattern of D (major), G, D, D, so it's clear to see that the overall symphony is in D major.

But when we get to Beethoven 5, we switch to C major at the end, yet we still consider it a symphony in C minor. But C major and minor are parallel keys, so this isn't too much of a stretch. Mahler 2 takes it the next step: the first movement is in C minor, but it switches to the relative key of E♭ major by the final movement; but we still say Mahler 2 is in C minor!

Mahler 5 then took it all the way: a symphony whose first movement is in C♯ minor ends with a movement in D major. Yet every score and recording will tell you it's a symphony in C♯ minor.

In short, it seems that we tend to privilege the final key when analyzing single movements. But in multi-movement works, we tend to use the key of the first movement to determine the key of the entire work. (And typically we find the key of the first movement by looking at its ending.)

And a comment below reminded me of another important distinction: that of ending in major compared to ending on major. When a piece ends in major, it's been in major for some time. This is different than a minor piece ending on a major chord on account of a Picardy third.

  • "I don't have firm data for this, but my experience as a musician (and conversations with other musicians) tells me that the general belief is that a composition's final key is its overall key." Not so. Particularly when the work is titled as being in a minor key. See the examples in my answer below. – Laurence Payne Jan 24 '18 at 12:50
  • @LaurencePayne But don't your answers follow the multi-movement stipulation that I placed in the latter half of my answer?: "But in multi-movement works, we tend to use the key of the first movement to determine the key of the entire work. (And typically we find the key of the first movement by looking at its ending.)" – Richard Jan 24 '18 at 13:01
  • If you derive the key of the first movement from its ending, a whole lot (but not all) of pieces that composer and publisher have agreed to label as 'in X minor' will have to be re-named! – Laurence Payne Jan 24 '18 at 13:05
  • Perhaps so. (Though I personally think there's a difference between endings with a Picardy third and endings where the last 50 measures are all in major.) I just wanted to point out that the examples you listed all follow my stipulations given above. – Richard Jan 24 '18 at 13:09
  • I think this is a great answer, and the best answer so far, but I am reluctant to mark it as the accepted answer as, like the other answers, it is primarily opinion-based. The consensus seems to be that the final substantial key of the first substantial section/movement determines the overall key, but I would like to see this confirmed with a quote from a definitive source. – HullCityFan852 Feb 11 '18 at 18:26

Yes, multi-movement works tend to be named for the key of the first substantial section. And it's very common for a work that starts off in a minor key to end up in either the tonic major or relative major. In fact, in the context of the 'classical' period - which could be defined as that when key relationships were paramount in defining the structure of a musical work - it would almost be unusual for it NOT to.

The Wikipedia quote:

The music [of the 4th movement] resounds in C major, an unusual choice by the composer as a symphony that begins in C minor is expected to finish in that key.

is rubbish. Let's look at a few symphonies 'in E minor'. Dvorak 9 (the 'New World') ends in E major. Tchaikovsky 5 ends in E major. Brahms 4 ends in E minor - it CAN happen :-) Rachmaninoff 2 ends in E major. I could go on... We could compare this reluctance to end in a minor key to the Tierce de Picardie, popular since about 1500.

  • 1
    I wonder if the Wikipedia editor meant that Beethoven 5 was unusual for its time? Dvorak, Tchaikovsky, Brahms, and Rachmaninoff all came after Papa B. – Richard Jan 24 '18 at 2:30
  • @Richard: Haydn has at least four minor-key symphonies that end in the parallel major (26, 45, 83, and 95). – Micah Jan 24 '18 at 4:38

Here's my guess for how finding overall keys works:

Multi-Movement Works

The name is probably the definitive, canon indicator for the key the multi-movement work is in. For example, Chopin's Piano Sonata No. 3 in B Minor, Op. 58 is in B minor, despite one of its movements being in B major and another in E flat major. Heck, the last movement ends in B major (and starts in B minor)! Another example is Mendelssohn's Symphony No. 4 in A Major, Op. 90, "Italian" is in A major, despite one of its movements being in D minor and its last movement being in A minor. Yet another example is Beethoven's Piano Sonata No. 27 in E Minor, Op. 90, which is in E minor despite being only 2 movements long--the 1st movement is in E minor and the last is in E major.

Once you've found enough multi-movement works with names that reveal their keys, you can sketch out some rough rules for determining the overall key of a multi-movement work. I generally agree with your proposed factors, except that I agree more strongly with the first key of the final movement being a major factor. This is because minor-key works often end in the tonic major.

I'll have to provide some caveats--some multi-movement works have no overall key, and some movements have no overall key. Your rough rules must be flexible enough to accommodate this. For example, whatever the key of Bartok's fifth string quartet is, it hasn't been widely published.

Single Movements

From listening to a LOT of music, here are my rough factors for finding what overall key a piece has, if any:

  • The key of the first motive
  • The key of the last motive
  • The most often used key
  • The percentage of the piece that is atonal
  • All keys that are playing simultaneously (in the case of polytonal music)
  • The key in the piece that has the closest relationships with all other keys in the piece (dominant/subdominant/tonic major are closer relationships than chromatic mediant, for example)
  • The assigned key of other pieces with the same form as this piece (e.g. marches tend to be assigned their 1st key as their overall key, even though they often don't end in the same key as they start with)

You can also figure out the overall keys of contemporary music with these guidelines.

Gary P. Gilroy's Heart of the City should be a fairly straightforward example--it starts in F minor, it ends in F minor, and it's in F minor the most often, so its overall key is F minor. It doesn't matter much that some of its phrases are in F major or A flat major.

A less straightforward example is C-R-O-W-N-E-D from Kirby's Return to Dream Land. It also starts in F minor and ends in F minor. However, portions of it are also in A flat minor, E minor, and C minor, and it quite possibly ends up in A flat minor the most often. Despite C minor arguably having the closest relationships to all the other keys (it's the dominant of F minor and chromatic mediants of all the other keys, while F minor has a very loose relationship with E minor), I'd still say the overall key of this piece is F minor. This piece still begins and ends in that key, after all.

For me at least, overall keys get more tenuous the higher percentage of the piece is atonal. While The Defender from Shovel Knight is in D minor overall (it arguably starts in D minor and is in D minor the most often), large portions of that piece are chromatic to the point of atonality, and a phrase of it is even in B minor.

Eventually, pieces stop having overall keys. Half or more of Ginastera's Danza del gaucho matrero, Op. 2, No. 3 from his Danzas Argentinas is atonal, and the rest is desperately trying to cling onto a C major-like tonic. Some of it is actually in C major (with an interlude in A flat major), but I'd hesitate to say this piece has any overall key.

By the way, given these tips, can you help me find the overall key (if any) of Mind in a PROGRAM from Kirby: Planet Robobot? Portions of it are in C minor, F# minor, B minor, D minor, G# minor, C# minor, and E minor, and whatever key the intro is in, it probably isn't any of those. But I cannot determine which of those keys is the overall one. Even though the largest percentage of this piece is in E minor, that's its ending, non-starting key. I'm starting to give up on determining whether this piece has an overall key. Does this piece have an overall key?


I agree that specifically the /first/ key of the opening movement, or the key of its first subject (after any introduction) is a deciding factor. In some multi-movement works, even the opening movements begin in a minor key and end in the tonic major. (And not just with a Tierce de Picardie.) For example:

C-V. Alkan. Grand duo concertante (Violin Sonata), op.21. (in f#)

Haydn. String quartet 59 ("Rider"), op.74 no.3 (in g)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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