diff --git a/crates/bevy_pbr/Cargo.toml b/crates/bevy_pbr/Cargo.toml
index d75be9f3c9d07..dc5b35a97b997 100644
--- a/crates/bevy_pbr/Cargo.toml
+++ b/crates/bevy_pbr/Cargo.toml
@@ -43,7 +43,7 @@ thiserror = { version = "1", optional = true }
 bitflags = "2.3"
 fixedbitset = "0.5"
 # direct dependency required for derive macro
-bytemuck = { version = "1", features = ["derive"] }
+bytemuck = { version = "1", features = ["derive", "must_cast"] }
 radsort = "0.1"
 smallvec = "1.6"
 serde = { version = "1", features = ["derive", "rc"] }
diff --git a/crates/bevy_pbr/src/meshlet/gpu_scene.rs b/crates/bevy_pbr/src/meshlet/gpu_scene.rs
index 492179dc2bcfe..9944f853d3f85 100644
--- a/crates/bevy_pbr/src/meshlet/gpu_scene.rs
+++ b/crates/bevy_pbr/src/meshlet/gpu_scene.rs
@@ -174,7 +174,7 @@ pub fn queue_material_meshlet_meshes<M: Material>(
 }
 
 // TODO: Try using Queue::write_buffer_with() in queue_meshlet_mesh_upload() to reduce copies
-fn upload_storage_buffer<T: ShaderSize + bytemuck::Pod>(
+fn upload_storage_buffer<T: ShaderSize + bytemuck::NoUninit>(
     buffer: &mut StorageBuffer<Vec<T>>,
     render_device: &RenderDevice,
     render_queue: &RenderQueue,
@@ -187,7 +187,7 @@ fn upload_storage_buffer<T: ShaderSize + bytemuck::Pod>(
 
     if capacity >= size {
         let inner = inner.unwrap();
-        let bytes = bytemuck::cast_slice(buffer.get().as_slice());
+        let bytes = bytemuck::must_cast_slice(buffer.get().as_slice());
         render_queue.write_buffer(inner, 0, bytes);
     } else {
         buffer.write_buffer(render_device, render_queue);
diff --git a/crates/bevy_pbr/src/render/morph.rs b/crates/bevy_pbr/src/render/morph.rs
index cdc9c599ea24b..b53cf62f2f8e9 100644
--- a/crates/bevy_pbr/src/render/morph.rs
+++ b/crates/bevy_pbr/src/render/morph.rs
@@ -11,7 +11,7 @@ use bevy_render::{
     view::ViewVisibility,
     Extract,
 };
-use bytemuck::Pod;
+use bytemuck::NoUninit;
 
 #[derive(Component)]
 pub struct MorphIndex {
@@ -54,7 +54,7 @@ const fn can_align(step: usize, target: usize) -> bool {
 const WGPU_MIN_ALIGN: usize = 256;
 
 /// Align a [`BufferVec`] to `N` bytes by padding the end with `T::default()` values.
-fn add_to_alignment<T: Pod + Default>(buffer: &mut BufferVec<T>) {
+fn add_to_alignment<T: NoUninit + Default>(buffer: &mut BufferVec<T>) {
     let n = WGPU_MIN_ALIGN;
     let t_size = mem::size_of::<T>();
     if !can_align(n, t_size) {
diff --git a/crates/bevy_render/Cargo.toml b/crates/bevy_render/Cargo.toml
index cbc25b9146b56..cec4260c60b27 100644
--- a/crates/bevy_render/Cargo.toml
+++ b/crates/bevy_render/Cargo.toml
@@ -80,7 +80,7 @@ wgpu = { version = "0.19.3", default-features = false, features = [
 naga = { version = "0.19", features = ["wgsl-in"] }
 serde = { version = "1", features = ["derive"] }
 bitflags = { version = "2.3", features = ["serde"] }
-bytemuck = { version = "1.5", features = ["derive"] }
+bytemuck = { version = "1.5", features = ["derive", "must_cast"] }
 downcast-rs = "1.2.0"
 thread_local = "1.1"
 thiserror = "1.0"
diff --git a/crates/bevy_render/src/render_resource/buffer_vec.rs b/crates/bevy_render/src/render_resource/buffer_vec.rs
index 26656c58ed983..f330d52dea50a 100644
--- a/crates/bevy_render/src/render_resource/buffer_vec.rs
+++ b/crates/bevy_render/src/render_resource/buffer_vec.rs
@@ -2,14 +2,14 @@ use crate::{
     render_resource::Buffer,
     renderer::{RenderDevice, RenderQueue},
 };
-use bytemuck::{cast_slice, Pod};
+use bytemuck::{must_cast_slice, NoUninit};
 use wgpu::BufferUsages;
 
 /// A structure for storing raw bytes that have already been properly formatted
 /// for use by the GPU.
 ///
 /// "Properly formatted" means that item data already meets the alignment and padding
-/// requirements for how it will be used on the GPU. The item type must implement [`Pod`]
+/// requirements for how it will be used on the GPU. The item type must implement [`NoUninit`]
 /// for its data representation to be directly copyable.
 ///
 /// Index, vertex, and instance-rate vertex buffers have no alignment nor padding requirements and
@@ -28,7 +28,7 @@ use wgpu::BufferUsages;
 /// * [`GpuArrayBuffer`](crate::render_resource::GpuArrayBuffer)
 /// * [`BufferVec`]
 /// * [`Texture`](crate::render_resource::Texture)
-pub struct BufferVec<T: Pod> {
+pub struct BufferVec<T: NoUninit> {
     values: Vec<T>,
     buffer: Option<Buffer>,
     capacity: usize,
@@ -38,7 +38,7 @@ pub struct BufferVec<T: Pod> {
     label_changed: bool,
 }
 
-impl<T: Pod> BufferVec<T> {
+impl<T: NoUninit> BufferVec<T> {
     pub const fn new(buffer_usage: BufferUsages) -> Self {
         Self {
             values: Vec::new(),
@@ -132,7 +132,7 @@ impl<T: Pod> BufferVec<T> {
         self.reserve(self.values.len(), device);
         if let Some(buffer) = &self.buffer {
             let range = 0..self.item_size * self.values.len();
-            let bytes: &[u8] = cast_slice(&self.values);
+            let bytes: &[u8] = must_cast_slice(&self.values);
             queue.write_buffer(buffer, 0, &bytes[range]);
         }
     }
@@ -154,7 +154,7 @@ impl<T: Pod> BufferVec<T> {
     }
 }
 
-impl<T: Pod> Extend<T> for BufferVec<T> {
+impl<T: NoUninit> Extend<T> for BufferVec<T> {
     #[inline]
     fn extend<I: IntoIterator<Item = T>>(&mut self, iter: I) {
         self.values.extend(iter);