Two decks, one subject. Maybe you downloaded a shared deck and made your own cards alongside it. Maybe a friend sent you their deck and you want the best of both. Merging them sounds like a five-minute task. It is not.
The manual way
1. Import the second deck. Double-click the .apkg or use File > Import. Anki brings it in as its own deck, sitting next to yours. Nothing is merged yet.
2. Move the cards over. Open the browser, select all cards in the imported deck, right-click, Change Deck, and move them into your main deck. Now everything lives in one place.
3. Hunt for duplicates. Go to Edit > Find Duplicates in the browser. This is where the manual method starts to fall apart: Anki only flags exact matches in the first field of a note. "What is the normal range of serum potassium?" and "Normal serum K+ range?" are the same fact in different words, and Find Duplicates will never catch them.
4. Clean up the note types. The imported deck probably brought its own note types and card templates. Your collection now has two "Basic" variants and a custom template with broken styling. Merging note types is a separate manual job, and getting it wrong breaks card rendering.
5. Fix the scheduling. If both decks were studied in the same profile, review history survives the move. If the second deck came from another device or profile, its scheduling does not transfer cleanly, and you end up with mature cards reset to new.
Steps 3 through 5 are where the hours go. A 3,000-card merge can eat an evening, and the reworded duplicates you missed will quietly double your daily reviews for months.
The better way: merge as a comparison problem
The merge is really a matching problem: which cards in deck B test facts already covered by deck A, which are genuinely new, and which conflict. SlideToAnki's deck tools do the matching for you, semantically rather than by exact text, so reworded duplicates get caught too.
Upload both .apkg files and pick Merge. Cards from both decks are kept, duplicates are removed, and anywhere the two decks disagree about the same fact you get a side-by-side conflict review: keep A, keep B, or keep both. You download one merged .apkg at the end.
What you skip: the import, the browser moves, the exact-match duplicate hunt that misses the real duplicates, and the note-type cleanup. The merge you would have done by hand if exact-text matching were enough.
When not to merge
Merging is not always the answer. If the two decks cover different subjects, keep them separate — mixed-subject decks make filtered decks and custom study sessions harder to aim. If one deck is strictly a subset of the other, you want the overlap operation instead: keep only the facts both decks cover. Merge when both decks have material worth keeping and you want one deck to rule the subject.