diff --git a/lib/Horde/String.php b/lib/Horde/String.php
index 2e1268c..eaea434 100644
--- a/lib/Horde/String.php
+++ b/lib/Horde/String.php
@@ -183,7 +183,10 @@ protected static function _convertCharset($input, $from, $to)
         if (class_exists('UConverter')) {
             $attemptedMethods[] = 'UConverter';
             try {
-                $conv = new UConverter($to, $from);
+                $conv = new UConverter(
+                    CharacterSets::toUConverter($to),
+                    CharacterSets::toUConverter($from)
+                );
                 $out = $conv->convert($input);
                 if ($out !== false && $out !== '') {
                     return $out;
diff --git a/src/CharacterSets.php b/src/CharacterSets.php
index f018f7e..160537f 100644
--- a/src/CharacterSets.php
+++ b/src/CharacterSets.php
@@ -35,6 +35,24 @@ class CharacterSets
         'unknown-8bit' => 'iso-8859-1',
     ];
 
+    /**
+     * Map charset aliases to unambiguous ICU canonical names for UConverter.
+     *
+     * PHP's UConverter emits "Ambiguous encoding specified" warnings for names
+     * that map to multiple ICU converters (e.g. windows-1258 → ibm-5354 or
+     * cp1258). Use the canonical name PHP would pick anyway.
+     *
+     * @see https://www.php.net/manual/en/class.uconverter.php
+     */
+    private static array $uconverterMap = [
+        'big5-hkscs' => 'ibm-1375_P100-2008',
+        'shift_jis' => 'ibm-943_P15A-2003',
+        'tis-620' => 'windows-874-2000',
+        'windows-1258' => 'cp1258',
+        'windows-936' => 'windows-936-2000',
+        'windows-950' => 'windows-950-2000',
+    ];
+
     /**
      * Normalize a character set identifier to a canonical name.
      *
@@ -65,4 +83,19 @@ public static function toMbstring(string $identifier): string
         // TODO: Check against mb_list_encoding
         return self::normalize($identifier);
     }
+
+    /**
+     * Convert charset identifier to an unambiguous UConverter/ICU name.
+     *
+     * @param string $identifier  The charset identifier.
+     *
+     * @return string  The UConverter-compatible charset name.
+     */
+    public static function toUConverter(string $identifier): string
+    {
+        $identifier = self::normalize($identifier);
+        $lower = strtolower($identifier);
+
+        return self::$uconverterMap[$lower] ?? $identifier;
+    }
 }
diff --git a/src/HordeString.php b/src/HordeString.php
index 71ebaf7..3dc9387 100644
--- a/src/HordeString.php
+++ b/src/HordeString.php
@@ -199,7 +199,10 @@ protected static function _convertCharset($input, $from, $to)
         if (class_exists('UConverter')) {
             $attemptedMethods[] = 'UConverter';
             try {
-                $conv = new UConverter($to, $from);
+                $conv = new UConverter(
+                    CharacterSets::toUConverter($to),
+                    CharacterSets::toUConverter($from)
+                );
                 $out = $conv->convert($input);
                 if ($out !== false && $out !== '') {
                     return $out;
diff --git a/test/CharacterSetsTest.php b/test/CharacterSetsTest.php
index e55c969..99cb2d3 100644
--- a/test/CharacterSetsTest.php
+++ b/test/CharacterSetsTest.php
@@ -78,4 +78,16 @@ public function testToMbstringPreservesOtherCharsets(): void
         // Non-normalized charsets should pass through
         $this->assertEquals('iso-8859-1', CharacterSets::toMbstring('iso-8859-1'));
     }
+
+    public function testToUConverterWindows1258(): void
+    {
+        $this->assertEquals('cp1258', CharacterSets::toUConverter('windows-1258'));
+        $this->assertEquals('cp1258', CharacterSets::toUConverter('Windows-1258'));
+    }
+
+    public function testToUConverterPreservesUnambiguousCharsets(): void
+    {
+        $this->assertEquals('iso-8859-1', CharacterSets::toUConverter('iso-8859-1'));
+        $this->assertEquals('utf-8', CharacterSets::toUConverter('utf8mb4'));
+    }
 }