diff --git a/apps/openmw/mwgui/spellmodel.cpp b/apps/openmw/mwgui/spellmodel.cpp
index ba661093f..d1db582d2 100644
--- a/apps/openmw/mwgui/spellmodel.cpp
+++ b/apps/openmw/mwgui/spellmodel.cpp
@@ -52,19 +52,21 @@ namespace MWGui
         const MWWorld::ESMStore &esmStore =
             MWBase::Environment::get().getWorld()->getStore();
 
+        std::string filter = Misc::StringUtils::lowerCaseUtf8(mFilter);
+
         for (MWMechanics::Spells::TIterator it = spells.begin(); it != spells.end(); ++it)
         {
             const ESM::Spell* spell = it->first;
             if (spell->mData.mType != ESM::Spell::ST_Power && spell->mData.mType != ESM::Spell::ST_Spell)
                 continue;
 
-            std::string name = spell->mName;
+            std::string name = Misc::StringUtils::lowerCaseUtf8(spell->mName);
 
-            if (name.find(mFilter) == std::string::npos)
+            if (name.find(filter) == std::string::npos)
                 continue;
 
             Spell newSpell;
-            newSpell.mName = name;
+            newSpell.mName = spell->mName;
             if (spell->mData.mType == ESM::Spell::ST_Spell)
             {
                 newSpell.mType = Spell::Type_Spell;
@@ -98,9 +100,9 @@ namespace MWGui
             if (enchant->mData.mType != ESM::Enchantment::WhenUsed && enchant->mData.mType != ESM::Enchantment::CastOnce)
                 continue;
 
-            std::string name = item.getClass().getName(item);
+            std::string name = Misc::StringUtils::lowerCaseUtf8(item.getClass().getName(item));
 
-            if (name.find(mFilter) == std::string::npos)
+            if (name.find(filter) == std::string::npos)
                 continue;
 
             Spell newSpell;
diff --git a/apps/openmw/mwgui/spellwindow.cpp b/apps/openmw/mwgui/spellwindow.cpp
index 10b4598ee..3fe171e4e 100644
--- a/apps/openmw/mwgui/spellwindow.cpp
+++ b/apps/openmw/mwgui/spellwindow.cpp
@@ -217,7 +217,7 @@ namespace MWGui
         if (stats.isParalyzed() || stats.getKnockedDown() || stats.isDead() || stats.getHitRecovery())
             return;
 
-        mSpellView->setModel(new SpellModel(MWMechanics::getPlayer(), mFilterEdit->getCaption()));
+        mSpellView->setModel(new SpellModel(MWMechanics::getPlayer(), ""));
 
         SpellModel::ModelIndex selected = 0;
         for (SpellModel::ModelIndex i = 0; i<int(mSpellView->getModel()->getItemCount()); ++i)
diff --git a/components/misc/stringops.hpp b/components/misc/stringops.hpp
index 9f4931d72..0fde1c96c 100644
--- a/components/misc/stringops.hpp
+++ b/components/misc/stringops.hpp
@@ -6,6 +6,8 @@
 #include <string>
 #include <algorithm>
 
+#include "utf8stream.hpp"
+
 namespace Misc
 {
 class StringUtils
@@ -56,6 +58,70 @@ public:
         };
     }
 
+    static Utf8Stream::UnicodeChar toLowerUtf8(Utf8Stream::UnicodeChar ch)
+    {
+        // Russian alphabet
+        if (ch >= 0x0410 && ch < 0x0430)
+            return ch += 0x20;
+
+        // Cyrillic IO character
+        if (ch == 0x0401)
+            return ch += 0x50;
+
+        // Latin alphabet
+        if (ch >= 0x41 && ch < 0x60)
+            return ch += 0x20;
+
+        // Deutch characters
+        if (ch == 0xc4 || ch == 0xd6 || ch == 0xdc)
+            return ch += 0x20;
+        if (ch == 0x1e9e)
+            return 0xdf;
+
+        // TODO: probably we will need to support characters from other languages
+
+        return ch;
+    }
+
+    static std::string lowerCaseUtf8(const std::string str)
+    {
+        if (str.empty())
+            return str;
+
+        // Decode string as utf8 characters, convert to lower case and pack them to string
+        std::string out;
+        Utf8Stream stream (str.c_str());
+        while (!stream.eof ())
+        {
+            Utf8Stream::UnicodeChar character = toLowerUtf8(stream.peek());
+
+            if (character <= 0x7f)
+                out.append(1, static_cast<char>(character));
+            else if (character <= 0x7ff)
+            {
+                out.append(1, static_cast<char>(0xc0 | ((character >> 6) & 0x1f)));
+                out.append(1, static_cast<char>(0x80 | (character & 0x3f)));
+            }
+            else if (character <= 0xffff)
+            {
+                out.append(1, static_cast<char>(0xe0 | ((character >> 12) & 0x0f)));
+                out.append(1, static_cast<char>(0x80 | ((character >> 6) & 0x3f)));
+                out.append(1, static_cast<char>(0x80 | (character & 0x3f)));
+            }
+            else
+            {
+                out.append(1, static_cast<char>(0xf0 | ((character >> 18) & 0x07)));
+                out.append(1, static_cast<char>(0x80 | ((character >> 12) & 0x3f)));
+                out.append(1, static_cast<char>(0x80 | ((character >> 6) & 0x3f)));
+                out.append(1, static_cast<char>(0x80 | (character & 0x3f)));
+            }
+
+            stream.consume();
+        }
+
+        return out;
+    }
+
     static bool ciLess(const std::string &x, const std::string &y) {
         return std::lexicographical_compare(x.begin(), x.end(), y.begin(), y.end(), ci());
     }