diff --git a/changelog.md b/changelog.md
new file mode 100644
index 00000000..7ec07261
--- /dev/null
+++ b/changelog.md
@@ -0,0 +1,79 @@
+# Changelog
+All notable changes to this project will be documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/), and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## [Unreleased] - 2019-04-29
+### Added
+### Changed
+### Fixed
+
+## [v0.5.0](https://github.com/psu-libraries/cho/releases/tag/v0.5.0) - 2019-01-11
+### Added
+- Agent records for creators, contributors and other controlled fields. Agents can be added via a form in the GUI or in bulk via CSV.
+- Subform for agents on metadata edit forms allowing value-pairs for related metadata, e.g. MARC relators for creators.
+- Data dictionary linked-field type for controlled vocabularies
+### Changed
+- Confirmation process for deleting collections and works
+- Thumbnails displayed in search results and on file set landing page
+- Facets include result count
+### Fixed
+- Tests generate correct type of random content
+
+## [v0.4.0](https://github.com/psu-libraries/cho/releases/tag/v0.4.0) - 2018-10-26
+### Added
+- Matomo web analytics in basic implementation
+- Footer with linked tag of the deployed release in production
+- Dates can be stored as ETDF dates, work ticketed for indexing and display improvements
+- Export collection metadata, including fileset metadata as CSV
+- Page for file sets
+
+### Changed
+- Edit form has been standardized to look and operate in a standardized and repeatable way
+- Batch creation of works (CSV) now includes fileset metadata, batch update still does not
+- Configure facets with the data dictionary
+- Files with machine-extractable text are indexed (not OCR), text stored as new file in fileset
+- Filenames are indexed
+
+### Fixed
+- Both archival and library collections now show up in the create dropdown menu.
+- Improved error messaging during creation workflow
+- Removed files and file sets from manage works batch selection
+
+## [v0.3.0](https://github.com/psu-libraries/cho/releases/tag/v0.3.0) - 2018-09-08
+### Added
+- Batch creation of works (CSV) with files; files are bagged and stagged on the CHO web server.
+- Files are organized into file sets and classified by usage (preservation, preservation redacted, service, access, thumb)
+- Files are characterized by FITS by output is a blob, not currently parsed
+
+### Changed
+- Accessibility improvements
+
+### Fixed
+- Bug fixes
+
+## [v0.2.0](https://github.com/psu-libraries/cho/releases/tag/v0.2.0) - 2018-06-04
+### Added
+- Files can be attached to works and stored, but not downloaded
+- Batch creation of works by uploading a CSV with metadata, no files
+- Batch update of work metadata by uploading a CSV
+- Delete works, individually and in batch
+- Facets for collections
+- Atomic indexing
+
+## [v0.1.0](https://github.com/psu-libraries/cho/releases/tag/v0.1.0) - 2017-10-27
+### Added
+- Rails web server
+- Blacklight Solr discovery interface for indexing, searching, and browsing of resources and collections
+- Valkyrie for storing metadata and files in Samvera
+- Forms for creating and describing archival and library collections and multiple work types
+- Pages for collections and works
+- Metadata defined by a data dictionary, with schemas for work types
+- Authentication using Penn States's Shibboleth
+
+## Code Diffs
+- Unreleased: https://github.com/psu-libraries/cho/compare/v0.5.0...HEAD
+- v0.4.0: https://github.com/psu-libraries/cho/compare/v0.4.0...v0.5.0
+- v0.4.0: https://github.com/psu-libraries/cho/compare/v0.3.0...v0.4.0
+- v0.3.0: https://github.com/psu-libraries/cho/compare/v0.2.0...v0.3.0
+- v0.2.0: https://github.com/psu-libraries/cho/compare/v0.1.0...v0.2.0