diff --git a/regex-automata/Cargo.toml b/regex-automata/Cargo.toml
index 40a0ebfb9e..3e0abcf985 100644
--- a/regex-automata/Cargo.toml
+++ b/regex-automata/Cargo.toml
@@ -87,6 +87,7 @@ aho-corasick = { version = "1.0.0", optional = true, default-features = false }
 log = { version = "0.4.14", optional = true }
 memchr = { version = "2.6.0", optional = true, default-features = false }
 regex-syntax = { path = "../regex-syntax", version = "0.8.2", optional = true, default-features = false }
+zerocopy = { version = "0.7.32", default-features = false, features = ["derive"] }
 
 [dev-dependencies]
 anyhow = "1.0.69"
diff --git a/regex-automata/src/dfa/accel.rs b/regex-automata/src/dfa/accel.rs
index c0ba18ea89..6120d92299 100644
--- a/regex-automata/src/dfa/accel.rs
+++ b/regex-automata/src/dfa/accel.rs
@@ -53,6 +53,8 @@
 #[cfg(feature = "dfa-build")]
 use alloc::{vec, vec::Vec};
 
+use zerocopy::{AsBytes, FromBytes};
+
 use crate::util::{
     int::Pointer,
     memchr,
@@ -208,15 +210,9 @@ impl<'a> Accels<&'a [AccelTy]> {
         wire::check_alignment::<AccelTy>(slice)?;
         let accel_tys = &slice[..accel_tys_bytes_len];
         slice = &slice[accel_tys_bytes_len..];
-        // SAFETY: We've checked the length and alignment above, and since
-        // slice is just bytes and AccelTy is just a u32, we can safely cast to
-        // a slice of &[AccelTy].
-        let accels = unsafe {
-            core::slice::from_raw_parts(
-                accel_tys.as_ptr().cast::<AccelTy>(),
-                accel_tys_len,
-            )
-        };
+        // PANICS: We've just checked the length and alignment, so this is
+        // guaranteed to succeed.
+        let accels = <AccelTy>::slice_from(accel_tys).unwrap();
         Ok((Accels { accels }, slice.as_ptr().as_usize() - slice_start))
     }
 }
@@ -235,15 +231,7 @@ impl<A: AsRef<[AccelTy]>> Accels<A> {
 
     /// Return the bytes representing the serialization of the accelerators.
     pub fn as_bytes(&self) -> &[u8] {
-        let accels = self.accels.as_ref();
-        // SAFETY: This is safe because accels is a just a slice of AccelTy,
-        // and u8 always has a smaller alignment.
-        unsafe {
-            core::slice::from_raw_parts(
-                accels.as_ptr().cast::<u8>(),
-                accels.len() * ACCEL_TY_SIZE,
-            )
-        }
+        self.accels.as_ref().as_bytes()
     }
 
     /// Returns the memory usage, in bytes, of these accelerators.
diff --git a/regex-automata/src/util/primitives.rs b/regex-automata/src/util/primitives.rs
index 5c5d187b0e..2c5929d641 100644
--- a/regex-automata/src/util/primitives.rs
+++ b/regex-automata/src/util/primitives.rs
@@ -34,6 +34,8 @@ use core::num::NonZeroUsize;
 #[cfg(feature = "alloc")]
 use alloc::vec::Vec;
 
+use zerocopy::{AsBytes, FromBytes, FromZeroes};
+
 use crate::util::int::{Usize, U16, U32, U64};
 
 /// A `usize` that can never be `usize::MAX`.
@@ -138,7 +140,18 @@ impl core::fmt::Debug for NonMaxUsize {
 /// an invalid value can be done in entirely safe code. This may in turn result
 /// in panics or silent logical errors.
 #[derive(
-    Clone, Copy, Debug, Default, Eq, Hash, PartialEq, PartialOrd, Ord,
+    Clone,
+    Copy,
+    Debug,
+    Default,
+    Eq,
+    Hash,
+    PartialEq,
+    PartialOrd,
+    Ord,
+    FromZeroes,
+    FromBytes,
+    AsBytes,
 )]
 #[repr(transparent)]
 pub struct SmallIndex(u32);
@@ -746,7 +759,19 @@ pub struct PatternID(SmallIndex);
 ///
 /// See the [`SmallIndex`] type for more information about what it means for
 /// a state ID to be a "small index."
-#[derive(Clone, Copy, Default, Eq, Hash, PartialEq, PartialOrd, Ord)]
+#[derive(
+    Clone,
+    Copy,
+    Default,
+    Eq,
+    Hash,
+    PartialEq,
+    PartialOrd,
+    Ord,
+    FromZeroes,
+    FromBytes,
+    AsBytes,
+)]
 #[repr(transparent)]
 pub struct StateID(SmallIndex);
 
diff --git a/regex-automata/src/util/wire.rs b/regex-automata/src/util/wire.rs
index b1351c7e91..ffca87f8c1 100644
--- a/regex-automata/src/util/wire.rs
+++ b/regex-automata/src/util/wire.rs
@@ -45,6 +45,7 @@ use core::{cmp, mem::size_of};
 
 #[cfg(feature = "alloc")]
 use alloc::{vec, vec::Vec};
+use zerocopy::{AsBytes, FromBytes};
 
 use crate::util::{
     int::Pointer,
@@ -268,32 +269,16 @@ impl core::fmt::Display for DeserializeError {
 /// Safely converts a `&[u32]` to `&[StateID]` with zero cost.
 #[cfg_attr(feature = "perf-inline", inline(always))]
 pub(crate) fn u32s_to_state_ids(slice: &[u32]) -> &[StateID] {
-    // SAFETY: This is safe because StateID is defined to have the same memory
-    // representation as a u32 (it is repr(transparent)). While not every u32
-    // is a "valid" StateID, callers are not permitted to rely on the validity
-    // of StateIDs for memory safety. It can only lead to logical errors. (This
-    // is why StateID::new_unchecked is safe.)
-    unsafe {
-        core::slice::from_raw_parts(
-            slice.as_ptr().cast::<StateID>(),
-            slice.len(),
-        )
-    }
+    // PANICS: This is guaranteed to succeed since `u32` and `StateID` have the
+    // same size and alignment.
+    StateID::slice_from(slice.as_bytes()).unwrap()
 }
 
 /// Safely converts a `&mut [u32]` to `&mut [StateID]` with zero cost.
 pub(crate) fn u32s_to_state_ids_mut(slice: &mut [u32]) -> &mut [StateID] {
-    // SAFETY: This is safe because StateID is defined to have the same memory
-    // representation as a u32 (it is repr(transparent)). While not every u32
-    // is a "valid" StateID, callers are not permitted to rely on the validity
-    // of StateIDs for memory safety. It can only lead to logical errors. (This
-    // is why StateID::new_unchecked is safe.)
-    unsafe {
-        core::slice::from_raw_parts_mut(
-            slice.as_mut_ptr().cast::<StateID>(),
-            slice.len(),
-        )
-    }
+    // PANICS: This is guaranteed to succeed since `u32` and `StateID` have the
+    // same size and alignment.
+    StateID::mut_slice_from(slice.as_bytes_mut()).unwrap()
 }
 
 /// Safely converts a `&[u32]` to `&[PatternID]` with zero cost.