: How to know whether to write accidentals as sharps or flats? I am writing a bassline for a song in Em and there is a passing note A# or Bb in the approach notes. How do I write this note? Since the key signature has one sharp,

If it really IS a passing note, 'sharpen it going up, flatten it going down' is a good guideline. This has nothing to do with the key signature.

A and B below involve a different set of decisions than C and D.

---

It really depends on your harmonic structure. It depends firstly on the scale you are in and then what you are trying to do.

When you have sharps in the key signature, you'll most likely use sharps as accidentals.

You'll choose your accidental depending on where you want to move afterwards. The case usually is sharp when you move up, flat when you move down. So, if you are in E minor (F# in the key signature) and you want to have a chromatic movement from A to B, you'll have A -> A# -> B; similarly if you want to move chromatically from B down to A, you'll have B -> Bb -> A.

---

I assume that this question arises because you're composing on a fixed-pitch instrument, tuned in 12-tone equal temperament, without distinction between enharmonic notes. Probably a piano.

For convenience, I shall denote the 12 notes of the chromatic scale with the integers modulo 12, using the convention that 0=C, 2=D, 4=E, 5=F, 7=G, 9=A, and 11=B; with 1, 3, 6, 8, and 10 being black keys.

The E (natural) minor scale consists of the notes E (4), F? (6), G (7), A (9), B (11), C (0), and D (2). Your bassline includes note 10 and you need to decide whether to notate it as A? or B? (or maybe even C).

Or more generally, how to denote the 5 chromatic notes that are not part of the diatonic E-minor scale.

Is note 1 C?, D?, or B?
Is note 3 D?, E?, or F?
Is note 5 E?, F, or G?
Is note 8 G? or A??
Is note 10 A?, B?, or C?

Approach #1 : Match the key signature

You already have a sharp in your diatonic scale, so just use sharps for the 5 chromatic notes: C?, D?, E?, G?, and A?.

For note 5, you could choose to use F instead of E?, but since you'd have to write an accidental (?) anyway to unsharpen the F?, there's no notational advantage in doing so.

Approach #2 : Tertian harmony

The popular types of chords in Western music are all built out of thirds: M3 (4 semitones) and m3 (3 semitones). There are 3 possible ways to reach note 10 from the E-minor scale and thirds:

F? (6) + M3 (4) = A? (10)
G (7) + m3 (3) = B? (10)
D (2) - M3 (4) = B? (10)

So if, for example, you intend this note to be the middle of a F?-major chord (F? + A? + C?), spell it as A?. If, OTOH, it's part of a G-minor chord (G + B?
+ D), spell it as B?.

Approach #3 : Circle of Fifths

Assume that notes closer together on the Circle of Fifths are more consonant or otherwise “better” when played together. (For now, let's not debate whether this is actually true.)

The Circle of Fifths order of notes, with the E-minor scale in brackets, is:

...-F?-C?-G?-D?-A?-E?-B?-F-[C-G-D-A-E-B-F?]-C?-G?-D?-A?-E?-B?-...

For note names as close to the bracket diatonic scale as possible:

Note 1 = C?, only 1 fifth outside the scale (vs. D? which is 5)
Note 5 = F?, also only 1 fifth outside the scale (vs. E? which is 5)
Note 8 = G?, 2 fifths outside the scale (vs. A? which is 4)
Note 10 = B?, 2 fifths outside the scale (vs. A? which is 4)
Note 3 is a tie between D? and E?, each 3 fifths outside the scale. Break the tie by noting that it's far more common to raise the seventh note in a minor scale (as in harmonic and melodic minor) than the lower the first. So D?.

The full E-minor chromatic scale in order is thus:

E, F?, F?, G, G?, A, B?, B, C, C?, D, D?, E

Or, using moveable-do solfège with the la-based (relative) minor convention:

la, ti? (te), ti, do, do? (di), re, mi? (me), mi, fa, fa? (fi), sol, sol? (si), la

---

Adding to Shev's answer, it also ought to reflect the harmonic structure of what chord is being represented. Although this is often ignored to make things easier to read. E.g. going from C>F, using C+ as the harmony, the G♯/A♭ should be the former, as the G has been sharpened; If going from F>C through Fm, the G♯/A♭ ought to be A♭, as that reflects the m3 of Fm.

Often in diminished chords, the fourth note of the chord is 'wrongly' spelled. E.g. in Co, it's actually B♭♭, but gets written as A. Technically wrong - but far easier to read, so is oft used. For more examples, see Mark Levine's books.

EDIT: re-visiting the question, we cannot answer properly, as we are not told which note/s come before and after the A#/Bb. Can the OP enlighten us?

But, the broad answer is keep to sharps in a sharp key, flats in a flat key. The reader is already wearing a matching hat.

---

It is strange to me that the emphasis of answers to this question are
on the theoretical "correctness" of notation. Yet, what is truly most
important is your intent when writing something down. Once notated
parts are performed, the resulting sounds are completely unaffected by the way in which they were notated.
Does a Bb sound different from an A#?...which leads to the question: Why should I notate something one way or another? The answer is actually quite simple! What is clear? What is concise? If you read it, is it confusing? When you get someone else to read it, does it make sense to them? Theoretically I could write a piece in E# major but how would that change the way it sounds compared to writing it in F? Theory is only a way to bridge our individual experiences of sound. It can be useful... Just don't forget that music doesn't actually exist, and that whatever achieves your
intent is what really matters.!

---