How do I import a PDF into Zotero?

Last reviewed: October 1, 2026 · Markdown version

Drag the PDF from your file manager into the Zotero window. Zotero copies it into its own storage, tries to work out what the paper is, and builds a full reference item with the PDF attached. When that lookup fails you are left with a bare "standalone" PDF, which you can fix in a minute. The guidance below is from Zotero's own documentation and covers the desktop app.

Drag it in, or use New Item

Zotero's metadata guide gives both routes: "simply drag your existing PDFs into your Zotero library or use the “Store Copy of File” or “Link to File” options from the add new item menu (green plus sign)." With the default settings, "Zotero will automatically retrieve metadata for each PDF, create an appropriate parent item, and rename the associated file based on the metadata"; those automatic steps can be switched off in the General pane of the preferences (Retrieve PDF Metadata — Zotero Documentation, read 1 October 2026).

Where you drop it matters. Per the file guide, files dropped onto an existing item in the centre pane become child items of it, while files dropped onto a collection, into empty space or between items "are added as standalone items" (Adding Files to your Zotero Library — Zotero Documentation, read 1 October 2026). To attach a PDF to a reference you already have, select the item and use the paperclip "Add Attachment" button.

Stored copy or linked file?

This is the one decision worth thinking about before importing a large folder.

Stored files are the default. The file "is copied to the Zotero data directory, so you may wish to delete the original to avoid confusion", Zotero manages it, and with file syncing turned on it travels to your other devices.
Linked files leave the PDF where it is. They "are not synced, nor are they deleted if the attachment item is deleted in Zotero", cannot be used in a group library, and "are also not supported in the Zotero iOS app". To link while dragging, hold Ctrl + Shift (Windows/Linux) or Cmd + Option (Mac) as you drop.

Zotero's own advice is unambiguous: "We strongly recommend using stored files for the most seamless experience" (Adding Files to your Zotero Library, read 1 October 2026). If you change your mind later, Tools → Manage Attachments converts linked files to stored ones.

When Zotero can't identify the PDF

The lookup "should retrieve high-quality metadata for most academic PDFs", but for other documents "you shouldn’t expect that — anything can be distributed as a PDF, but that doesn’t mean there’s any standard metadata available for it." If no match is found, right-click the PDF, choose Create Parent Item, and enter a DOI or ISBN; failing that, click Manual Entry. And if a match is wrong, right-click the new parent and choose "Undo Retrieve Metadata" (Retrieve PDF Metadata, read 1 October 2026).

How the lookup works explains the most common failure. "The Zotero client sends the first few pages of text from the PDF to the web service", which matches it against Crossref and DOI and ISBN lookups; the same page notes the service "doesn’t require a Zotero account and doesn’t log any data about the content or results of searches." The page does not mention scanned PDFs. Our reading is that a scan with no text layer gives the service nothing to work from, so if a scanned article will not identify itself, check it for real text before typing the reference in by hand.

The faster route for new papers

Zotero itself says importing PDFs "is not the best way to add items to your library in general." Its recommendation is the Zotero Connector in your browser, used on the article's page rather than on the PDF: it "will usually save high-quality metadata and also automatically download the relevant PDF if you have access to it" (Adding Items to Zotero — Zotero Documentation, read 1 October 2026). Keep drag-and-drop for the PDFs you already have on disk.

Once your PDFs are in, searching inside them in Zotero has its own quirks, including a default limit on how many pages of each file get indexed.