diff --git a/CHANGELOG.md b/CHANGELOG.md
index b6f97fa..6fd23be 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -7,7 +7,9 @@ All notable changes to this project will be documented in this file.
 ### Added
 
 * Support for tags.
-    * Filter notes by their tags using `--tag "history, europe"`. To match notes associated with either tags, use a pipe `|` or `OR` (all caps), e.g. `--tag "inbox OR todo"`.
+    * Filter notes by their tags using `--tag "history, europe"`.
+        * To match notes associated with either tags, use a pipe `|` or `OR` (all caps), e.g. `--tag "inbox OR todo"`.
+        * If you want to exclude notes having a particular tag, prefix it with `-` or `NOT` (all caps), e.g. `--tag "NOT done"`
     * Many tag flavors are supported: `#hashtags`, `:colon:separated:tags:` and even Bear's [`#multi-word tags#`](https://blog.bear.app/2017/11/bear-tips-how-to-create-multi-word-tags/). If you prefer to use a YAML frontmatter, list your tags with the key `tags` or `keywords`.
 
 ### Changed
diff --git a/adapter/sqlite/note_dao.go b/adapter/sqlite/note_dao.go
index 34e1be6..0328a4e 100644
--- a/adapter/sqlite/note_dao.go
+++ b/adapter/sqlite/note_dao.go
@@ -442,11 +442,22 @@ func (d *NoteDAO) findRows(opts note.FinderOpts) (*sql.Rows, error) {
 				break
 			}
 			separatorRegex := regexp.MustCompile(`(\ OR\ )|\|`)
-			for _, tags := range filter {
-				tags := separatorRegex.Split(tags, -1)
+			for _, tagsArg := range filter {
+				tags := separatorRegex.Split(tagsArg, -1)
 
+				negate := false
 				globs := make([]string, 0)
 				for _, tag := range tags {
+					tag = strings.TrimSpace(tag)
+
+					if strings.HasPrefix(tag, "-") {
+						negate = true
+						tag = strings.TrimPrefix(tag, "-")
+					} else if strings.HasPrefix(tag, "NOT") {
+						negate = true
+						tag = strings.TrimPrefix(tag, "NOT")
+					}
+
 					tag = strings.TrimSpace(tag)
 					if len(tag) == 0 {
 						continue
@@ -455,13 +466,25 @@ func (d *NoteDAO) findRows(opts note.FinderOpts) (*sql.Rows, error) {
 					args = append(args, tag)
 				}
 
-				whereExprs = append(whereExprs, fmt.Sprintf(`n.id IN (
+				if len(globs) == 0 {
+					continue
+				}
+				if negate && len(globs) > 1 {
+					return nil, fmt.Errorf("cannot negate a tag in a OR group: %s", tagsArg)
+				}
+
+				expr := "n.id"
+				if negate {
+					expr += " NOT"
+				}
+				expr += fmt.Sprintf(` IN (
 SELECT note_id FROM notes_collections
  WHERE collection_id IN (SELECT id FROM collections t WHERE kind = '%s' AND (%s))
  )`,
 					note.CollectionKindTag,
 					strings.Join(globs, " OR "),
-				))
+				)
+				whereExprs = append(whereExprs, expr)
 			}
 
 		case note.ExcludePathFilter:
diff --git a/adapter/sqlite/note_dao_test.go b/adapter/sqlite/note_dao_test.go
index 2a0a006..c7fcd18 100644
--- a/adapter/sqlite/note_dao_test.go
+++ b/adapter/sqlite/note_dao_test.go
@@ -418,6 +418,9 @@ func TestNoteDAOFindTag(t *testing.T) {
 	test([]string{"fiction | adventure | fantasy"}, []string{"ref/test/b.md", "f39c8.md", "log/2021-01-03.md"})
 	test([]string{"fiction | history", "adventure"}, []string{"ref/test/b.md", "log/2021-01-03.md"})
 	test([]string{"fiction", "unknown"}, []string{})
+	test([]string{"-fiction"}, []string{"ref/test/b.md", "f39c8.md", "ref/test/a.md", "log/2021-02-04.md", "index.md", "log/2021-01-04.md"})
+	test([]string{"NOT   fiction"}, []string{"ref/test/b.md", "f39c8.md", "ref/test/a.md", "log/2021-02-04.md", "index.md", "log/2021-01-04.md"})
+	test([]string{"NOTfiction"}, []string{"ref/test/b.md", "f39c8.md", "ref/test/a.md", "log/2021-02-04.md", "index.md", "log/2021-01-04.md"})
 }
 
 func TestNoteDAOFindMatch(t *testing.T) {