diff --git a/linera-witty/src/runtime/memory.rs b/linera-witty/src/runtime/memory.rs
index fa54916e1f7..33e6e26d586 100644
--- a/linera-witty/src/runtime/memory.rs
+++ b/linera-witty/src/runtime/memory.rs
@@ -11,6 +11,10 @@ use crate::{Layout, WitType};
 use frunk::{hlist, hlist_pat, HList};
 use std::borrow::Cow;
 
+#[cfg(test)]
+#[path = "unit_tests/memory.rs"]
+mod tests;
+
 /// An address for a location in a guest WebAssembly module's memory.
 #[derive(Clone, Copy, Debug, Eq, PartialEq)]
 pub struct GuestPointer(pub(crate) u32);
@@ -18,25 +22,28 @@ pub struct GuestPointer(pub(crate) u32);
 impl GuestPointer {
     /// Returns a new address that's the current address advanced to add padding to ensure it's
     /// aligned to the `alignment` byte boundary.
-    pub fn aligned_at(&self, alignment: u32) -> Self {
+    pub const fn aligned_at(&self, alignment: u32) -> Self {
+        // The following computation is equivalent to:
+        // `(alignment - (self.0 % alignment)) % alignment`.
+        // Source: https://en.wikipedia.org/wiki/Data_structure_alignment#Computing_padding
         let padding = (-(self.0 as i32) & (alignment as i32 - 1)) as u32;
 
         GuestPointer(self.0 + padding)
     }
 
     /// Returns a new address that's the current address advanced to after the size of `T`.
-    pub fn after<T: WitType>(&self) -> Self {
+    pub const fn after<T: WitType>(&self) -> Self {
         GuestPointer(self.0 + T::SIZE)
     }
 
     /// Returns a new address that's the current address advanced to add padding to ensure it's
     /// aligned properly for `T`.
-    pub fn after_padding_for<T: WitType>(&self) -> Self {
+    pub const fn after_padding_for<T: WitType>(&self) -> Self {
         self.aligned_at(<T::Layout as Layout>::ALIGNMENT)
     }
 
     /// Returns the address of an element in a contiguous list of properly aligned `T` types.
-    pub fn index<T: WitType>(&self, index: u32) -> Self {
+    pub const fn index<T: WitType>(&self, index: u32) -> Self {
         let element_size = GuestPointer(T::SIZE).after_padding_for::<T>();
 
         GuestPointer(self.0 + index * element_size.0)
diff --git a/linera-witty/src/runtime/unit_tests/memory.rs b/linera-witty/src/runtime/unit_tests/memory.rs
new file mode 100644
index 00000000000..aefe838a6c9
--- /dev/null
+++ b/linera-witty/src/runtime/unit_tests/memory.rs
@@ -0,0 +1,25 @@
+// Copyright (c) Zefchain Labs, Inc.
+// SPDX-License-Identifier: Apache-2.0
+
+//! Unit tests for guest Wasm module memory manipulation.
+
+use super::GuestPointer;
+
+/// Test aligning memory addresses.
+///
+/// Check that the resulting address is aligned and that it never advances more than the alignment
+/// amount.
+#[test]
+fn align_guest_pointer() {
+    for alignment_bits in 0..3 {
+        let alignment = 1 << alignment_bits;
+        let alignment_mask = alignment - 1;
+
+        for start_offset in 0..32 {
+            let address = GuestPointer(start_offset).aligned_at(alignment);
+
+            assert_eq!(address.0 & alignment_mask, 0);
+            assert!(address.0 - start_offset < alignment);
+        }
+    }
+}
diff --git a/linera-witty/src/type_traits/implementations/frunk.rs b/linera-witty/src/type_traits/implementations/frunk.rs
index b57c4a2b049..e5e35b1ee12 100644
--- a/linera-witty/src/type_traits/implementations/frunk.rs
+++ b/linera-witty/src/type_traits/implementations/frunk.rs
@@ -66,11 +66,11 @@ impl WitStore for HNil {
 impl<Head, Tail> WitType for HCons<Head, Tail>
 where
     Head: WitType,
-    Tail: WitType,
+    Tail: WitType + SizeCalculation,
     Head::Layout: Add<Tail::Layout>,
     <Head::Layout as Add<Tail::Layout>>::Output: Layout,
 {
-    const SIZE: u32 = Head::SIZE + Tail::SIZE;
+    const SIZE: u32 = Self::SIZE_STARTING_AT_BYTE_BOUNDARIES[0];
 
     type Layout = <Head::Layout as Add<Tail::Layout>>::Output;
 }
@@ -78,7 +78,7 @@ where
 impl<Head, Tail> WitLoad for HCons<Head, Tail>
 where
     Head: WitLoad,
-    Tail: WitLoad,
+    Tail: WitLoad + SizeCalculation,
     Head::Layout: Add<Tail::Layout>,
     <Head::Layout as Add<Tail::Layout>>::Output: Layout,
     <Self::Layout as Layout>::Flat:
@@ -94,7 +94,12 @@ where
     {
         Ok(HCons {
             head: Head::load(memory, location)?,
-            tail: Tail::load(memory, location.after::<Head>())?,
+            tail: Tail::load(
+                memory,
+                location
+                    .after::<Head>()
+                    .after_padding_for::<Tail::FirstElement>(),
+            )?,
         })
     }
 
@@ -118,7 +123,7 @@ where
 impl<Head, Tail> WitStore for HCons<Head, Tail>
 where
     Head: WitStore,
-    Tail: WitStore,
+    Tail: WitStore + SizeCalculation,
     Head::Layout: Add<Tail::Layout>,
     <Head::Layout as Add<Tail::Layout>>::Output: Layout,
     <Head::Layout as Layout>::Flat: Add<<Tail::Layout as Layout>::Flat>,
@@ -136,7 +141,12 @@ where
         <Instance::Runtime as Runtime>::Memory: RuntimeMemory<Instance>,
     {
         self.head.store(memory, location)?;
-        self.tail.store(memory, location.after::<Head>())?;
+        self.tail.store(
+            memory,
+            location
+                .after::<Head>()
+                .after_padding_for::<Tail::FirstElement>(),
+        )?;
 
         Ok(())
     }
@@ -155,3 +165,58 @@ where
         Ok(head_layout + tail_layout)
     }
 }
+
+/// Helper trait used to calculate the size of a heterogeneous list considering internal alignment.
+///
+/// Assumes the maximum alignment necessary for any type is 8 bytes, which is the alignment for the
+/// largest flat types (`i64` and `f64`).
+trait SizeCalculation {
+    /// The size of the list considering the current size calculation starts at different offsets
+    /// inside an 8-byte window.
+    const SIZE_STARTING_AT_BYTE_BOUNDARIES: [u32; 8];
+
+    /// The type of the first element of the list, used to determine the current necessary
+    /// alignment.
+    type FirstElement: WitType;
+}
+
+impl SizeCalculation for HNil {
+    const SIZE_STARTING_AT_BYTE_BOUNDARIES: [u32; 8] = [0; 8];
+
+    type FirstElement = ();
+}
+
+/// Unrolls a `for`-like loop so that it runs in a `const` context.
+macro_rules! unroll_for {
+    ($binding:ident in [ $($elements:expr),* $(,)? ] $body:tt) => {
+        $(
+            let $binding = $elements;
+            $body
+        )*
+    };
+}
+
+impl<Head, Tail> SizeCalculation for HCons<Head, Tail>
+where
+    Head: WitType,
+    Tail: SizeCalculation,
+{
+    const SIZE_STARTING_AT_BYTE_BOUNDARIES: [u32; 8] = {
+        let mut size_at_boundaries = [0; 8];
+
+        unroll_for!(boundary_offset in [0, 1, 2, 3, 4, 5, 6, 7] {
+            let memory_location = GuestPointer(boundary_offset)
+                .after_padding_for::<Head>()
+                .after::<Head>();
+
+            let tail_size = Tail::SIZE_STARTING_AT_BYTE_BOUNDARIES[memory_location.0 as usize % 8];
+
+            size_at_boundaries[boundary_offset as usize] =
+                memory_location.0 - boundary_offset + tail_size;
+        });
+
+        size_at_boundaries
+    };
+
+    type FirstElement = Head;
+}
diff --git a/linera-witty/src/type_traits/implementations/mod.rs b/linera-witty/src/type_traits/implementations/mod.rs
index 5e7785d5048..bd729252a2c 100644
--- a/linera-witty/src/type_traits/implementations/mod.rs
+++ b/linera-witty/src/type_traits/implementations/mod.rs
@@ -6,3 +6,5 @@
 mod custom_types;
 mod frunk;
 mod std;
+#[cfg(test)]
+mod tests;
diff --git a/linera-witty/src/type_traits/implementations/tests.rs b/linera-witty/src/type_traits/implementations/tests.rs
new file mode 100644
index 00000000000..57cb32c36b2
--- /dev/null
+++ b/linera-witty/src/type_traits/implementations/tests.rs
@@ -0,0 +1,106 @@
+// Copyright (c) Zefchain Labs, Inc.
+// SPDX-License-Identifier: Apache-2.0
+
+//! Unit tests for implementations of the custom traits for existing types.
+
+use crate::{FakeInstance, InstanceWithMemory, Layout, WitLoad, WitStore};
+use frunk::hlist;
+use std::fmt::Debug;
+
+/// Test roundtrip of a heterogeneous list that doesn't need any internal padding.
+#[test]
+fn hlist_without_padding() {
+    let input = hlist![
+        0x1011_1213_1415_1617_1819_1a1b_1c1d_1e1f_u128,
+        0x2021_2223_2425_2627_i64,
+        0x3031_3233_u32,
+        0x4041_i16,
+        true,
+    ];
+
+    test_memory_roundtrip(
+        input,
+        &[
+            0x1f, 0x1e, 0x1d, 0x1c, 0x1b, 0x1a, 0x19, 0x18, 0x17, 0x16, 0x15, 0x14, 0x13, 0x12,
+            0x11, 0x10, 0x27, 0x26, 0x25, 0x24, 0x23, 0x22, 0x21, 0x20, 0x33, 0x32, 0x31, 0x30,
+            0x41, 0x40, 0x01,
+        ],
+    );
+    test_flattening_roundtrip(
+        input,
+        hlist![
+            0x1819_1a1b_1c1d_1e1f_i64,
+            0x1011_1213_1415_1617_i64,
+            0x2021_2223_2425_2627_i64,
+            0x3031_3233_i32,
+            0x0000_4041_i32,
+            0x0000_0001_i32,
+        ],
+    );
+}
+
+/// Test roundtrip of a heterogeneous list that needs internal padding between some of its elements.
+#[test]
+fn hlist_with_padding() {
+    let input = hlist![
+        true,
+        0x1011_i16,
+        0x2021_u16,
+        0x3031_3233_u32,
+        0x4041_4243_4445_4647_i64,
+    ];
+
+    test_memory_roundtrip(
+        input,
+        &[
+            0x01, 0, 0x11, 0x10, 0x21, 0x20, 0, 0, 0x33, 0x32, 0x31, 0x30, 0, 0, 0, 0, 0x47, 0x46,
+            0x45, 0x44, 0x43, 0x42, 0x41, 0x40,
+        ],
+    );
+    test_flattening_roundtrip(
+        input,
+        hlist![
+            0x0000_0001_i32,
+            0x0000_1011_i32,
+            0x0000_2021_i32,
+            0x3031_3233_i32,
+            0x4041_4243_4445_4647_i64,
+        ],
+    );
+}
+
+/// Test storing an instance of `T` to memory, checking that the `memory_data` bytes are correctly
+/// written, and check that the instance can be loaded from those bytes.
+fn test_memory_roundtrip<T>(input: T, memory_data: &[u8])
+where
+    T: Debug + Eq + WitLoad + WitStore,
+{
+    let mut instance = FakeInstance::default();
+    let mut memory = instance.memory().unwrap();
+    let length = memory_data.len() as u32;
+
+    assert_eq!(length, T::SIZE);
+
+    let address = memory.allocate(length).unwrap();
+
+    input.store(&mut memory, address).unwrap();
+
+    assert_eq!(memory.read(address, length).unwrap(), memory_data);
+    assert_eq!(T::load(&memory, address).unwrap(), input);
+}
+
+/// Test lowering an instance of `T`, checking that the resulting flat layout matches the expected
+/// `flat_layout`, and check that the instance can be lifted from that flat layout.
+fn test_flattening_roundtrip<T>(input: T, flat_layout: <T::Layout as Layout>::Flat)
+where
+    T: Debug + Eq + WitLoad + WitStore,
+    <T::Layout as Layout>::Flat: Debug + Eq,
+{
+    let mut instance = FakeInstance::default();
+    let mut memory = instance.memory().unwrap();
+
+    let lowered_layout = input.lower(&mut memory).unwrap();
+
+    assert_eq!(lowered_layout, flat_layout);
+    assert_eq!(T::lift_from(lowered_layout, &memory).unwrap(), input);
+}
diff --git a/linera-witty/tests/common/types.rs b/linera-witty/tests/common/types.rs
index 5f714948fbf..01bd9631b32 100644
--- a/linera-witty/tests/common/types.rs
+++ b/linera-witty/tests/common/types.rs
@@ -13,7 +13,7 @@ pub struct SimpleWrapper(pub bool);
 #[derive(Clone, Copy, Debug, Eq, PartialEq, WitType, WitLoad, WitStore)]
 pub struct TupleWithoutPadding(pub u64, pub i32, pub i16);
 
-/// A tuple struct that requires internal padding in its memory layout between all of its fields.
+/// A tuple struct that requires internal padding in its memory layout between two of its fields.
 #[derive(Clone, Copy, Debug, Eq, PartialEq, WitType, WitLoad, WitStore)]
 pub struct TupleWithPadding(pub u16, pub u32, pub i64);