This page walks you through migrating an existing music library into Lidarr. It is intended for users who already have audio files on disk and want Lidarr to take over management of that library, rather than for users starting with an empty root folder.
If you are setting Lidarr up for the first time and do not already have files, use the Quick Start instead. If you are not sure whether Lidarr is a good fit for the library you have, read Concepts — Is Lidarr right for your library? first; importing a library Lidarr cannot model will waste a lot of time.
The automated import is scheduled, cannot be stopped once started, and can take hours on large libraries. Do not add a
Root Folderthat contains existing files until you have read this page in full.
Lidarr uses an automated process to scan a Root Folder, match the files it finds against MusicBrainz metadata, and add the resulting Release Artists and Releases to your library. For that to work, a few things need to be true about the library you're importing.
Releases (albums, EPs, singles, broadcasts) attributed to a single Release Artist. Loose collections, multi-artist folders that aren't compilations, and libraries of DJ mixes or beat packs will not import well. See Concepts.Release Artist → Release → tracks. A single folder with thousands of files in it will not work.Release.If one or more of these isn't true today, you have two options: fix it before import, or leave the problematic portion of your library outside Lidarr and import only the parts that conform. Both are legitimate choices; see Concepts.
See the FAQ entries How does Lidarr work? and How does Lidarr find releases? for the scheduling model and how searches are triggered. Lidarr does not continuously crawl indexers.
For the automated import to succeed, files should be structured and tagged before you point Lidarr at them. The amount of work here depends entirely on how clean your library already is.
The recommended structure is:
<Root Folder>/<Release Artist>/<Release>/<Track>
For example:
/music/Bob Dylan/Blood on the Tracks/01 - Tangled Up in Blue.flac
/music/Bob Dylan/Blood on the Tracks/02 - Simple Twist of Fate.flac
This layout, combined with proper tags, gives recognition rates of 95% or better in both Lidarr and most downstream media players. Flat layouts (one folder containing many releases), grouped-by-year layouts, or layouts that split discs into separate folders will all cause mismatches or import failures.
Disc-subfolder layouts (.../Release/CD1/..., .../Release/CD2/...) are a common trap: Lidarr expects all tracks for a release to live under the release folder, so consolidate or flatten disc subfolders before importing.
Tagging is where most import failures come from. Lidarr leans on MusicBrainz identifiers where they exist, and falls back to text matching on artist, album, and track fields where they don't — so the quality and consistency of your tags directly determines the quality of the import.
Recommended tagging tools:
Most of these tools will rename folders and restructure files at the same time as retagging, which means a single pass can fix both the structure and the tags. Associating MusicBrainz Release IDs with each file during this pass is the single biggest thing you can do to improve import accuracy.
Detailed use of Picard, Beets, and similar tools is beyond the scope of this wiki. See those projects' own documentation.
Once your files are structured and tagged, check the following before pointing Lidarr at them. A failed import of a large library is expensive to recover from.
Releases with very large track or disc counts dominate import time. As a rule of thumb, pull releases over ~25 discs or ~250 tracks out of the root folder and import them manually after the bulk pass.Release Artist with thousands of Releases in MusicBrainz will slow the scan substantially, even if only a handful of them are on your disk. This is rarely a blocker but is worth knowing about if the scan seems stuck on a particular artist.Releases, heavily dependent on disk speed, network latency to the metadata service, and system performance. Plan for an overnight run on large libraries.With files prepared and the pre-import checks done, the actual import is triggered by adding the library folder as a Root Folder. Lidarr starts scanning as soon as the folder is saved.
Releases (e.g. all, future, latest, none). This applies to every artist and release created by the import; you can change individual artists afterwards.Releases are cutoff-met and what Lidarr will search for.Release types (studio albums, EPs, singles, etc.) are visible on each imported artist.The three asterisked fields above become the defaults for every artist the import creates. You can change them on a per-artist basis later, but it is much less work to pick the right defaults up front than to batch-update hundreds of artists afterwards. In particular, Metadata Profile has the biggest downstream effect — a too-narrow profile will silently hide Releases that exist on disk.
Release (using MusicBrainz IDs where present, falling back to artist/album text matching).Release Artist and Release.Release Artist to the library, then adds the matched Releases, then links the existing files to their Release tracks.The scan cannot be stopped partway through. If you realise something is wrong (wrong path, wrong metadata profile, etc.) you can either let the scan finish and clean up afterwards, or stop Lidarr and remove the root folder before restarting — but there is no in-app cancel button.
Release Artist as it's added.Once the scan finishes, plan on a review pass. Even a well-prepared library will produce a small number of mismatches, and finding them early is cheaper than discovering them weeks later when Lidarr downloads the "missing" half of a mis-linked Release.
Releases. Usually means the metadata profile filtered everything out; widen the profile or change it for that artist.Release than the one you own. Use the Manual Import or Search for MusicBrainz ID workflow to re-link.folder.jpg, *.nfo, stray .txt files) are harmless. Duplicates and pre-import backup copies are worth cleaning up now so they don't confuse later searches.Most import problems are MusicBrainz data problems or tag problems, not Lidarr bugs. See the FAQ for specific failure modes — including artists that can't be added, releases missing from MusicBrainz, and what to do when an import refuses to match a file you're sure you've tagged correctly.
If a Release or Release Artist you expect is genuinely missing from MusicBrainz, you can add or correct it upstream: see How To Contribute. Once upstream propagation completes (hours to days), refresh the affected artist inside Lidarr and rescan.
Release and Artist model that drives import behavior