Most advice about studying mathematics is useless, or actively harmful.
It is usually written by people who succeeded in math environments despite their study habits, not because of them. They will tell you to form a study group. They will tell you to ask questions in lecture. They will tell you not to worry about understanding everything, and to trust it will click later.
This is mostly wrong. Here is what I think actually works.
1. Master the definitions before anything else
A math textbook is a structure. Every theorem rests on earlier theorems, which rest on axioms, which rest on definitions. If the definitions are fuzzy in your head, nothing built on top of them will hold.
Spend much longer on definitions than you think you should. Read each one three times. Write it out in your own words. Build examples. Build counter-examples. Notice what the definition excludes, not just what it includes.
When a later proof confuses you, the problem is almost always a definition that wasn't really solid. Go back.
2. Don't move on when you don't understand
The hardest rule. Also the most important.
There is enormous social pressure in math to seem like you get things fast. Lectures move on. Classmates nod. Professors say "this is obvious." Most of this is theater. If you don't understand something, stop, and actually work it out.
A week spent properly understanding one chapter is worth a month of skimming four.
3. Read the textbook, slowly, with a pen
Lectures are too fast. They skip steps. They wave hands. They depend on context the lecturer has and you don't. The textbook is the actual source.
Read it slowly. Do not continue to the next line until the current line makes sense. Write as you read. Rewrite proofs in your own words. Try to reproduce them with the book closed. If you cannot, you haven't understood the proof. You've understood that you have seen the proof.
4. Do the exercises. All of them.
There is no substitute for this. Mathematics is a skill before it is a body of knowledge, and skills come from doing the thing repeatedly.
Patterns that seem arbitrary in the abstract become obvious after the tenth exercise. Tools that feel clunky become automatic. Don't skip the easy ones. Don't skip the hard ones. Especially don't skip the ones that look like busywork. The busywork is where the intuition lives.
5. Work alone, mostly
Study groups are oversold. They are useful in specific, limited cases: working through a problem you've already tried alone for hours, or explaining something to someone else as a way of testing your own understanding.
They are not useful as your primary mode of study. Technical work requires sustained concentration that other people disrupt. If a group seems to be helping you study, check whether you are actually learning, or whether the group is functioning as mutual reassurance.
6. Take the subject literally
When a textbook says "let X be a group," it does not mean "something kind of like a group." It means exactly what the definition of a group says, no more and no less. Every word in a mathematical statement is load-bearing.
This is one of the pleasures of the subject. You can trust what is written. Other fields require you to interpret, to guess the writer's intent, to read between lines. Mathematics is one of the few places where the words mean exactly what they say. Use this.
7. Stay in the confusion longer than feels reasonable
Learning real mathematics is not like learning most things. You will feel like you understand nothing, for months. Then something will quietly click, and a large region of the subject will become obvious in a single afternoon.
Being confused for long stretches is not evidence you are bad at math. It is what learning math feels like. Everyone goes through it. The people who succeed are the ones who stay in the confusion long enough for the click to happen.
8. Sleep
Proofs consolidate overnight. Something you cannot see at 11 p.m. will often be obvious at 8 a.m.
This is not metaphor. Sleep reorganizes memory in ways that help with structural understanding. Work long enough to be genuinely stuck, then stop.
The overall shape
If there is a single principle behind all of this, it is: trust the subject, not the performance around it.
Mathematics rewards the people who actually try to understand, slowly, from the bottom. It does not reward people who look like they understand. These are different skills. The first is learnable. The second is not worth learning.