diff --git a/CHANGELOG.md b/CHANGELOG.md
index 1f573c09d1..f144829109 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -201,6 +201,12 @@ By @brodycj in [#6924](https://github.com/gfx-rs/wgpu/pull/6924).
 
 - Improve efficiency of dropping read-only buffer mappings. By @kpreid in [#7007](https://github.com/gfx-rs/wgpu/pull/7007).
 
+### Performance
+
+#### Naga
+
+- Replace `unicode-xid` with `unicode-ident`. By @CrazyboyQCD in [#7135](https://github.com/gfx-rs/wgpu/pull/7135)
+
 ### Documentation
 
 - Improved documentation around pipeline caches and `TextureBlitter`. By @DJMcNab in [#6978](https://github.com/gfx-rs/wgpu/pull/6978) and [#7003](https://github.com/gfx-rs/wgpu/pull/7003).
diff --git a/Cargo.lock b/Cargo.lock
index ac0a215b6d..327137d009 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -2353,7 +2353,7 @@ dependencies = [
  "strum 0.26.3",
  "termcolor",
  "thiserror 2.0.11",
- "unicode-xid",
+ "unicode-ident",
 ]
 
 [[package]]
diff --git a/naga/Cargo.toml b/naga/Cargo.toml
index 9cd92d1318..4458405bbf 100644
--- a/naga/Cargo.toml
+++ b/naga/Cargo.toml
@@ -54,7 +54,7 @@ spv-out = ["dep:spirv"]
 wgsl-in = [
     "dep:hexf-parse",
     "dep:strum",
-    "dep:unicode-xid",
+    "dep:unicode-ident",
     "indexmap/std",
     "compact",
 ]
@@ -97,7 +97,7 @@ serde = { version = "1.0.217", features = [
 petgraph = { version = "0.7", optional = true }
 pp-rs = { version = "0.2.1", optional = true }
 hexf-parse = { version = "0.2.1", optional = true }
-unicode-xid = { version = "0.2.6", optional = true }
+unicode-ident = { version = "1.0", optional = true }
 
 [build-dependencies]
 cfg_aliases.workspace = true
diff --git a/naga/src/front/wgsl/parse/lexer.rs b/naga/src/front/wgsl/parse/lexer.rs
index 0b0b6edebc..d55720972e 100644
--- a/naga/src/front/wgsl/parse/lexer.rs
+++ b/naga/src/front/wgsl/parse/lexer.rs
@@ -191,12 +191,12 @@ const fn is_blankspace(c: char) -> bool {
 
 /// Returns whether or not a char is a word start (Unicode XID_Start + '_')
 fn is_word_start(c: char) -> bool {
-    c == '_' || unicode_xid::UnicodeXID::is_xid_start(c)
+    c == '_' || unicode_ident::is_xid_start(c)
 }
 
 /// Returns whether or not a char is a word part (Unicode XID_Continue)
 fn is_word_part(c: char) -> bool {
-    unicode_xid::UnicodeXID::is_xid_continue(c)
+    unicode_ident::is_xid_continue(c)
 }
 
 #[derive(Clone)]