diff --git a/src/import/html_extract_text.cpp b/src/import/html_extract_text.cpp
index 0ff74da7..908ad0a8 100644
--- a/src/import/html_extract_text.cpp
+++ b/src/import/html_extract_text.cpp
@@ -121,6 +121,12 @@ namespace lily_of_the_valley
     //------------------------------------------------------------------
     void html_extract_text::parse_raw_text(const wchar_t* text, size_t textSize)
         {
+        /* Note about superscripts and subscripts.
+           Some pages apply this to entire paragraphs to make them appear in
+           a smaller font, which is what browsers then render them as.
+           There is no way to tell what the author's intent is, but if more than
+           four characters, then it's probably not really meant to be a super or subscript.*/
+        constexpr size_t maxSubscriptLength{ 4 };
         if (textSize > 0)
             {
             size_t currentStartPosition{ 0 };
@@ -344,14 +350,29 @@ namespace lily_of_the_valley
                         if (m_superscript_stack > 0)
                             {
                             // convert what we can along the way
-                            for (size_t i = 0; i < index; ++i)
-                                { add_character(string_util::to_superscript(text[i])); }
+                            if (index <= maxSubscriptLength)
+                                {
+                                for (size_t i = 0; i < index; ++i)
+                                    { add_character(string_util::to_superscript(text[i])); }
+                                }
+                            else
+                                {
+                                add_characters({ text, index });
+                                }
                             }
                         else if (m_subscript_stack > 0)
                             {
                             // convert what we can along the way
-                            for (size_t i = 0; i < index; ++i)
-                                { add_character(string_util::to_subscript(text[i])); }
+                            // (if really a subscript)
+                            if (index <= maxSubscriptLength)
+                                {
+                                for (size_t i = 0; i < index; ++i)
+                                    { add_character(string_util::to_subscript(text[i])); }
+                                }
+                            else
+                                {
+                                add_characters({ text, index });
+                                }
                             }
                         else
                             { add_characters(text, index); }
@@ -376,14 +397,29 @@ namespace lily_of_the_valley
                 if (m_superscript_stack > 0)
                     {
                     // convert what we can along the way
+                    if (textSize <= maxSubscriptLength)
+                        {
                     for (size_t i = 0; i < textSize; ++i)
                         { add_character(string_util::to_superscript(text[i])); }
+                        }
+                    else
+                        {
+                        add_characters({ text, textSize });
+                        }
                     }
                 else if (m_subscript_stack > 0)
                     {
                     // convert what we can along the way
-                    for (size_t i = 0; i < textSize; ++i)
-                        { add_character(string_util::to_subscript(text[i])); }
+                    // (if really a subscript)
+                    if (textSize <= maxSubscriptLength)
+                        {
+                        for (size_t i = 0; i < textSize; ++i)
+                            { add_character(string_util::to_subscript(text[i])); }
+                        }
+                    else
+                        {
+                        add_characters({ text, textSize });
+                        }
                     }
                 else
                     { add_characters(text, textSize); }
diff --git a/tests/htmlimporttests.cpp b/tests/htmlimporttests.cpp
index d3bf52f5..f49cfe9f 100644
--- a/tests/htmlimporttests.cpp
+++ b/tests/htmlimporttests.cpp
@@ -191,6 +191,76 @@ TEST_CASE("Str Chr Not Quoted", "[html import]")
         }
 	}
 
+TEST_CASE("HTML parser subscripts", "[html import]")
+    {
+        SECTION("Superscript")
+        {
+        html_extract_text filter_html;
+        const wchar_t* text = L"H<sup>2</sup>O<sup>37i</sup>";
+        std::wstring p = filter_html(text, std::wcslen(text), true, false);
+        CHECK(std::wstring(L"H²O³⁷ⁱ") == p);
+        text = L"H<sup>2</sup>O<sup>37Zi</sup>";
+        p = filter_html(text, std::wcslen(text), true, false);
+        CHECK(std::wstring(L"H²O³⁷Zⁱ") == p);
+        }
+    SECTION("Subscript")
+        {
+        html_extract_text filter_html;
+        const wchar_t* text = L"H<sub>2</sub>O<sub>37h</sub>";
+        std::wstring p = filter_html(text, std::wcslen(text), true, false);
+        CHECK(std::wstring(L"H₂O₃₇ₕ") == p);
+        text = L"H<sub>2</sub>O<sub>37Zh</sub>";
+        p = filter_html(text, std::wcslen(text), true, false);
+        CHECK(std::wstring(L"H₂O₃₇Zₕ") == p);
+        }
+    SECTION("Not really a script")
+        {
+        html_extract_text filter_html;
+        wchar_t* text = L"<sub>Hello22</sub> some text <sub>Hello2</sub>";
+        std::wstring p = filter_html(text, std::wcslen(text), true, false);
+        CHECK(std::wstring(L"Hello22 some text Hello2") == p);
+
+        text = L"<sup>Hello2</sup>";
+        p = filter_html(text, std::wcslen(text), true, false);
+        CHECK(std::wstring(L"Hello2") == p);
+        }
+    }
+
+TEST_CASE("HTML parser tags", "[html import]")
+    {
+    SECTION("Find Tag")
+        {
+        const wchar_t* text = L"body bgcolor=\'#FF0000\' color=\'#FF0000\'>there<br />world<br >!";
+        CHECK(html_extract_text::find_tag(text, L"bgcolor", 7, false) == text+5);
+        CHECK(html_extract_text::find_tag(text, L"BGCOLOR", 7, false) == text+5);
+        CHECK(html_extract_text::find_tag(text, L"color", 5, false) == text+23);
+        CHECK(html_extract_text::find_tag(text, L"width", 5, false) == nullptr);
+        CHECK(html_extract_text::find_tag(nullptr, L"width", 5, false) == nullptr);
+        CHECK(html_extract_text::find_tag(text, nullptr, 0, false) == nullptr);
+        CHECK(html_extract_text::find_tag(text, L"body", 4, false) == text);
+        }
+    SECTION("Find Tag 2")
+        {
+        const wchar_t* text = L"body style=\"color=#FF0000 width=250\">there<br />world<br >!";
+        CHECK(html_extract_text::find_tag(text, L"STYLE", 5, false) == text+5);
+        CHECK(html_extract_text::find_tag(text, L"color", 5, false) == nullptr);
+        CHECK(html_extract_text::find_tag(text, L"width", 5, false) == nullptr);
+        CHECK(html_extract_text::find_tag(nullptr, L"width", 5, false) == nullptr);
+        CHECK(html_extract_text::find_tag(text, nullptr, 0, false) == nullptr);
+        CHECK(html_extract_text::find_tag(text, L"body", 4, false) == text);
+        }
+    SECTION("Find Tag Quotable")
+        {
+        const wchar_t* text = L"body style=\"color=#FF0000 width=250\">there<br />world<br >!";
+        CHECK(html_extract_text::find_tag(text, L"STYLE", 5, true) == text+5);
+        CHECK(html_extract_text::find_tag(text, L"color", 5, true) == text+12);
+        CHECK(html_extract_text::find_tag(text, L"width", 5, true) == text+26);
+        CHECK(html_extract_text::find_tag(nullptr, L"width", 5, true) == nullptr);
+        CHECK(html_extract_text::find_tag(text, nullptr, 0, true) == nullptr);
+        CHECK(html_extract_text::find_tag(text, L"body", 4, true) == text);
+        }
+    }
+
 TEST_CASE("HTML Parser", "[html import]")
     {
     SECTION("Find Bookmark")
@@ -345,26 +415,6 @@ TEST_CASE("HTML Parser", "[html import]")
         p = filter_html(text, std::wcslen(text), true, false);
         CHECK(std::wstring(L"hello\n\n\n\nItem 1:\tThe definition\n\n") == p);
         }
-    SECTION("Superscript")
-        {
-        html_extract_text filter_html;
-        const wchar_t* text = L"H<sup>2</sup>O<sup>37i</sup>";
-        std::wstring p = filter_html(text, std::wcslen(text), true, false);
-        CHECK(std::wstring(L"H²O³⁷ⁱ") == p);
-        text = L"H<sup>2</sup>O<sup>37Zi</sup>";
-        p = filter_html(text, std::wcslen(text), true, false);
-        CHECK(std::wstring(L"H²O³⁷Zⁱ") == p);
-        }
-    SECTION("Subscript")
-        {
-        html_extract_text filter_html;
-        const wchar_t* text = L"H<sub>2</sub>O<sub>37h</sub>";
-        std::wstring p = filter_html(text, std::wcslen(text), true, false);
-        CHECK(std::wstring(L"H₂O₃₇ₕ") == p);
-        text = L"H<sub>2</sub>O<sub>37Zh</sub>";
-        p = filter_html(text, std::wcslen(text), true, false);
-        CHECK(std::wstring(L"H₂O₃₇Zₕ") == p);
-        }
     SECTION("Table")
         {
         html_extract_text filter_html;
@@ -462,37 +512,6 @@ TEST_CASE("HTML Parser", "[html import]")
         p = filter_html(text, std::wcslen(text), true, false);
         CHECK(std::wcscmp(p, L"A comprehensive (4 pages long) review of") == 0);
         }
-    SECTION("Find Tag")
-        {
-        const wchar_t* text = L"body bgcolor=\'#FF0000\' color=\'#FF0000\'>there<br />world<br >!";
-        CHECK(html_extract_text::find_tag(text, L"bgcolor", 7, false) == text+5);
-        CHECK(html_extract_text::find_tag(text, L"BGCOLOR", 7, false) == text+5);
-        CHECK(html_extract_text::find_tag(text, L"color", 5, false) == text+23);
-        CHECK(html_extract_text::find_tag(text, L"width", 5, false) == nullptr);
-        CHECK(html_extract_text::find_tag(nullptr, L"width", 5, false) == nullptr);
-        CHECK(html_extract_text::find_tag(text, nullptr, 0, false) == nullptr);
-        CHECK(html_extract_text::find_tag(text, L"body", 4, false) == text);
-        }
-    SECTION("Find Tag 2")
-        {
-        const wchar_t* text = L"body style=\"color=#FF0000 width=250\">there<br />world<br >!";
-        CHECK(html_extract_text::find_tag(text, L"STYLE", 5, false) == text+5);
-        CHECK(html_extract_text::find_tag(text, L"color", 5, false) == nullptr);
-        CHECK(html_extract_text::find_tag(text, L"width", 5, false) == nullptr);
-        CHECK(html_extract_text::find_tag(nullptr, L"width", 5, false) == nullptr);
-        CHECK(html_extract_text::find_tag(text, nullptr, 0, false) == nullptr);
-        CHECK(html_extract_text::find_tag(text, L"body", 4, false) == text);
-        }
-    SECTION("Find Tag Quotable")
-        {
-        const wchar_t* text = L"body style=\"color=#FF0000 width=250\">there<br />world<br >!";
-        CHECK(html_extract_text::find_tag(text, L"STYLE", 5, true) == text+5);
-        CHECK(html_extract_text::find_tag(text, L"color", 5, true) == text+12);
-        CHECK(html_extract_text::find_tag(text, L"width", 5, true) == text+26);
-        CHECK(html_extract_text::find_tag(nullptr, L"width", 5, true) == nullptr);
-        CHECK(html_extract_text::find_tag(text, nullptr, 0, true) == nullptr);
-        CHECK(html_extract_text::find_tag(text, L"body", 4, true) == text);
-        }
     SECTION("Read Element As String")
         {
         const wchar_t* text = L"<h1>My header</H1>";