diff --git a/CHANGELOG.md b/CHANGELOG.md
index b48ffbb..a99335e 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,3 +1,8 @@
+## 0.2.4
+
+- Deleted `go.dkinom.dev/baseconv/characters`
+- Added tests in `go.dkinom.dev/baseconv/internal/characters`
+
 ## 0.2.3
 
 - Retracted `go.dkinom.dev/baseconv/characters`
diff --git a/characters/doc.go b/characters/doc.go
deleted file mode 100644
index b703a42..0000000
--- a/characters/doc.go
+++ /dev/null
@@ -1 +0,0 @@
-package characters
diff --git a/characters/go.mod b/characters/go.mod
deleted file mode 100644
index 773751a..0000000
--- a/characters/go.mod
+++ /dev/null
@@ -1,5 +0,0 @@
-module go.dkinom.dev/baseconv/characters
-
-go 1.18
-
-retract [v0.0.0-00000000000000-000000000000, v0.2.3]
diff --git a/internal/characters/ascii_test.go b/internal/characters/ascii_test.go
new file mode 100644
index 0000000..d75cb7e
--- /dev/null
+++ b/internal/characters/ascii_test.go
@@ -0,0 +1,54 @@
+package characters
+
+import (
+	"testing"
+)
+
+func TestIsASCII(t *testing.T) {
+	t.Run("ASCII string", func(t *testing.T) {
+		if !isASCII("asdfgh") {
+			t.Fatal("'asdfgh' contains only ascii characters")
+		}
+	})
+
+	t.Run("Empty string", func(t *testing.T) {
+		if !isASCII("") {
+			t.Fatal()
+		}
+	})
+
+	t.Run("Non-ASCII string", func(t *testing.T) {
+		if isASCII("jklπoi") {
+			t.Fatal()
+		}
+	})
+}
+
+func TestASCIICharacterIterator(t *testing.T) {
+	t.Run("ASCII string", func(t *testing.T) {
+		str := "qwerty"
+		bytes := []byte(str)
+		itr := newASCII(str)
+
+		length := 0
+		for itr.Next() {
+			length++
+
+			if itr.Str() != string(bytes[length-1]) {
+				t.Fatal()
+			}
+		}
+
+		if length != 6 {
+			t.Fatal("'qwerty' has 6 characters")
+		}
+	})
+
+	t.Run("Empty string", func(t *testing.T) {
+		itr := newASCII("")
+
+		if itr.Next() {
+			t.Fatal()
+		}
+	})
+}
diff --git a/internal/characters/characters_test.go b/internal/characters/characters_test.go
new file mode 100644
index 0000000..0d4346f
--- /dev/null
+++ b/internal/characters/characters_test.go
@@ -0,0 +1,62 @@
+package characters
+
+import (
+	"testing"
+
+	"github.com/rivo/uniseg"
+)
+
+func TestCharacterIterator(t *testing.T) {
+	t.Run("ASCII string", func(t *testing.T) {
+		str := "zxcvbnm"
+		bytes := []byte(str)
+		itr := NewCharacters(str)
+
+		if _, ok := itr.(*ascii); !ok {
+			t.Fatal("expected ascii iterator")
+		}
+
+		length := 0
+		for itr.Next() {
+			length++
+
+			if itr.Str() != string(bytes[length-1]) {
+				t.Fatal()
+			}
+		}
+
+		if length != 7 {
+			t.Fatal("'zxcvbnm' has 7 characters")
+		}
+	})
+
+	t.Run("Empty string", func(t *testing.T) {
+		itr := NewCharacters("")
+
+		if _, ok := itr.(*ascii); !ok {
+			t.Fatal("expected ascii iterator")
+		}
+
+		if itr.Next() {
+			t.Fatal()
+		}
+	})
+
+	t.Run("Non-ASCII string", func(t *testing.T) {
+		str := "0️⃣1️⃣2️⃣3️⃣4️⃣5️⃣6️⃣7️⃣8️⃣9️⃣"
+		itr := NewCharacters(str)
+
+		if _, ok := itr.(*uniseg.Graphemes); !ok {
+			t.Fatal("expected graphemes iterator")
+		}
+
+		length := 0
+		for itr.Next() {
+			length++
+		}
+
+		if length != 10 {
+			t.Fatal("'0️⃣1️⃣2️⃣3️⃣4️⃣5️⃣6️⃣7️⃣8️⃣9️⃣' has 10 characters")
+		}
+	})
+}