diff --git a/MANIFEST b/MANIFEST
index 2e16bd3c..3c90c64f 100644
--- a/MANIFEST
+++ b/MANIFEST
@@ -53,6 +53,8 @@ t/51bind_type_guessing.t
 t/52comment.t
 t/53comment.t
 t/55utf8.t
+t/55utf8_errors.t
+t/55utf8_identifiers.t
 t/55utf8mb4.t
 t/56connattr.t
 t/57trackgtid.t
diff --git a/t/55utf8_errors.t b/t/55utf8_errors.t
new file mode 100644
index 00000000..5eb9d276
--- /dev/null
+++ b/t/55utf8_errors.t
@@ -0,0 +1,111 @@
+use strict;
+use warnings;
+
+use Test::More;
+use DBI;
+use Encode;
+
+use vars qw($test_dsn $test_user $test_password);
+use lib 't', '.';
+require "lib.pl";
+
+sub skip_rt_102404 {
+    skip "(Perl 5.13.1 and DBI 1.635) or DBI 1.639 is required due to bug RT 102404", $_[0] unless ($] >= 5.013001 and eval { DBI->VERSION(1.635) }) or eval { DBI->VERSION(1.639) };
+}
+
+my $dbh;
+eval {
+    $dbh = DBI->connect($test_dsn, $test_user, $test_password,
+                      { RaiseError => 1, PrintError => 1, AutoCommit => 0 });
+};
+if ($@) {
+    plan skip_all => "no database connection";
+}
+$dbh->disconnect();
+
+plan tests => 10 * 3;
+
+# All in internal Perl Unicode
+my $jpnErr = qr/\x{4ed8}\x{8fd1}.*\x{884c}\x{76ee}/; # Use \x{...} instead \N{U+...} due to Perl 5.12.0 bug
+
+foreach my $mysql_enable_utf8 (0, 1, 2) {
+    my %utf8_params = ();
+    if ($mysql_enable_utf8 == 1) {
+        $utf8_params{'mysql_enable_utf8'} = 1;
+        diag "Enabled mysql_enable_utf8.";
+    # XXX There are no utf8mb4 error characters
+    } elsif ($mysql_enable_utf8 == 2) {
+        $utf8_params{'mysql_enable_utf8mb4'} = 1;
+        diag "Enabled mysql_enable_utf8mb4.";
+    } else {
+        diag "Disabled mysql_enable_utf8.";
+    }
+    $dbh = DBI->connect($test_dsn, $test_user, $test_password,
+                       { RaiseError => 1, PrintError => 1, AutoCommit => 1, %utf8_params });
+
+    eval {
+        $dbh->do("SET lc_messages = 'ja_JP'");
+    } or do {
+        $dbh->disconnect();
+        plan skip_all => "Server lc_messages ja_JP are needed for this test";
+    };
+
+    my $sth;
+    my $warn;
+    my $dieerr;
+    my $dbierr;
+    my $failed;
+
+    $failed = 0;
+    $dieerr = undef;
+    $dbierr = undef;
+    $dbh->{HandleError} = sub { $dbierr = $_[0]; die $_[0]; };
+    eval {
+        $sth = $dbh->prepare("foo");
+        $sth->execute();
+        1;
+    } or do {
+        $dieerr = $@;
+        $failed = 1;
+    };
+    $dbh->{HandleError} = undef;
+
+    ok($failed, 'Execution of bad statement is failing (HandleError version).');
+    like(Encode::decode('UTF-8', $dbierr), $jpnErr, 'DBI error is in octets (HandleError version).'); # XXX
+    like(Encode::decode('UTF-8', $DBI::errstr), $jpnErr, 'DBI::errstr is in octets (HandleError version).'); # XXX 
+    like(Encode::decode('UTF-8', $dbh->errstr), $jpnErr, 'DBI handler errstr() method is in octets (HandleError version).'); # XXX
+
+    SKIP : {
+        skip_rt_102404 1;
+        like(Encode::decode('UTF-8', $dieerr), $jpnErr, 'Error from eval is in octets (HandleError version).');
+    }
+
+    $failed = 0;
+    $warn = undef;
+    $dieerr = undef;
+    $dbh->{PrintError} = 1;
+    $SIG{__WARN__} = sub { $warn = $_[0] };
+    eval {
+        $sth = $dbh->prepare("foo");
+        $sth->execute();
+        1;
+    } or do {
+        $dieerr = $@;
+        $failed = 1;
+    };
+    $dbh->{PrintError} = 0;
+    $SIG{__WARN__} = 'DEFAULT';
+
+    ok($failed, 'Execution of bad statement is failing (PrintError version).');
+    like(Encode::decode('UTF-8', $DBI::errstr), $jpnErr, 'DBI::errstr is in octets (PrintError version).'); # XXX
+    like(Encode::decode('UTF-8', $dbh->errstr), $jpnErr, 'DBI handler errstr() method is in octets (PrintError version).'); # XXX
+
+    SKIP : {
+        skip_rt_102404 2;
+        like(Encode::decode('UTF-8', $warn), $jpnErr, 'Warning is in octets (PrintError version).'); # XXX
+        like(Encode::decode('UTF-8', $dieerr), $jpnErr, 'Error from eval is in octets (PrintError version).'); # XXX
+    }
+
+    $dbh->disconnect();
+}
+done_testing;
diff --git a/t/55utf8_identifiers.t b/t/55utf8_identifiers.t
new file mode 100644
index 00000000..ed53e9b9
--- /dev/null
+++ b/t/55utf8_identifiers.t
@@ -0,0 +1,127 @@
+use strict;
+use warnings;
+
+use Test::More;
+use DBI;
+use Encode;
+
+use vars qw($test_dsn $test_user $test_password);
+use lib 't', '.';
+require "lib.pl";
+
+sub for_db {
+    my ($mysql_enable_utf8, $value) = @_; # Value is in internal Perl Unicode.
+
+    my $ret;
+    if ($mysql_enable_utf8 >= 1) {
+        $ret = $value;
+    } else {
+        $ret = Encode::encode('UTF-8', $value);
+    }
+
+    return $ret;
+}
+
+my $dbh;
+eval {
+    $dbh = DBI->connect($test_dsn, $test_user, $test_password,
+                       { RaiseError => 1, PrintError => 1, AutoCommit => 0 });
+};
+if ($@) {
+    plan skip_all => "no database connection";
+}
+$dbh->disconnect();
+
+plan tests => 12 * 3 + 12 * 2;
+
+# All in internal Perl Unicode
+my $jpnTable = "\N{U+8868}"; # Japanese table
+my $jpnColumn = "\N{U+6027}\N{U+5225}"; # Japanese column - word "gender"
+my $jpnData1 = "\N{U+5c71}\N{U+7530}\N{U+592a}\N{U+90ce}"; # Japanese data - person name
+my $jpnData2 = "\N{U+7537}"; # Japanese daya - word "male"
+my $chiTable = "\N{U+5927}\N{U+99AC}"; # Chinese table XXX MySQL doesn't support utf8mb4 in table names
+my $chiColumn = "\N{U+5C0F}\N{U+96EA}\N{U+4EBA}"; # Chinese column XXX MySQL doesn't support utf8mb4 in column names
+my $chiData1 = "\N{U+30001}"; # Chinese data
+my $chiData2 = "\N{U+30002}"; # Chinese data
+
+foreach my $mysql_enable_utf8 (0, 1, 2) {
+    my %utf8_params = ();
+    if ($mysql_enable_utf8 == 1) {
+        $utf8_params{'mysql_enable_utf8'} = 1;
+        diag "Enabled mysql_enable_utf8.";
+    } elsif ($mysql_enable_utf8 == 2) {
+        $utf8_params{'mysql_enable_utf8mb4'} = 1;
+        diag "Enabled mysql_enable_utf8mb4.";
+    } else {
+        diag "Disabled mysql_enable_utf8.";
+    }
+    $dbh = DBI->connect($test_dsn, $test_user, $test_password,
+                       { RaiseError => 1, PrintError => 1, AutoCommit => 1, %utf8_params });
+
+    my $jpnTable_db = for_db($mysql_enable_utf8, $jpnTable);
+    my $jpnColumn_db = for_db($mysql_enable_utf8, $jpnColumn);
+    my $jpnData1_db = for_db($mysql_enable_utf8, $jpnData1);
+    my $jpnData2_db = for_db($mysql_enable_utf8, $jpnData2);
+    my ($chiTable_db, $chiColumn_db, $chiData1_db, $chiData2_db);
+    if ($mysql_enable_utf8 == 0 || $mysql_enable_utf8 == 2) {
+        $chiTable_db = for_db($mysql_enable_utf8, $chiTable);
+        $chiColumn_db = for_db($mysql_enable_utf8, $chiColumn);
+        $chiData1_db = for_db($mysql_enable_utf8, $chiData1);
+        $chiData2_db = for_db($mysql_enable_utf8, $chiData2);
+    }
+
+    my $sth;
+    my $row;
+
+    ok($dbh->do("DROP TABLE IF EXISTS $jpnTable_db"), 'Drop table for Japanese testing.');
+    if ($mysql_enable_utf8 == 0 || $mysql_enable_utf8 == 2) {
+        ok($dbh->do("DROP TABLE IF EXISTS $chiTable_db"), 'Drop table for Chinese testings.');
+    }
+
+    ok($dbh->do(<<"END"
+CREATE TABLE IF NOT EXISTS $jpnTable_db (
+  name VARCHAR(20),
+  $jpnColumn_db CHAR(1)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8 COLLATE=utf8_bin
+END
+    ), 'Create temporay table with Japanese characters.');
+    if ($mysql_enable_utf8 == 0 || $mysql_enable_utf8 == 2) {
+      ok($dbh->do(<<"END"
+CREATE TABLE IF NOT EXISTS $chiTable_db (
+  name VARCHAR(20),
+  $chiColumn_db CHAR(1)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_bin
+END
+      ), 'Create temporay table with Chinese characters.');
+    }
+
+    ok($sth = $dbh->prepare("INSERT INTO $jpnTable_db (name, $jpnColumn_db) VALUES (?, ?)"), 'Prepare insert statement with Japanese values.');
+    ok($sth->execute($jpnData1_db, $jpnData2_db), 'Execute insert statement with Japanese values.');
+    if ($mysql_enable_utf8 == 0 || $mysql_enable_utf8 == 2) {
+        ok($sth = $dbh->prepare("INSERT INTO $chiTable_db (name, $chiColumn_db) VALUES (?, ?)"), 'Prepare insert statement with Chinese values.');
+        ok($sth->execute($chiData1_db, $chiData2_db), 'Execute insert statement with Chinese values.');
+    }
+
+    ok($sth = $dbh->prepare("SELECT * FROM $jpnTable_db"), 'Prepare select statement with Japanese values.');
+    ok($sth->execute(), 'Execute select statement with Japanese values.');
+    ok($row = $sth->fetchrow_hashref(), 'Fetch hashref with Japanese values.');
+    is($row->{name}, $jpnData1_db, "Japanese value.");
+    ok(!exists $row->{$jpnColumn}, 'Not exists Japanese key in internal Perl Unicode.'); # XXX
+    is($row->{Encode::encode('UTF-8', $jpnColumn)}, $jpnData2_db, 'Exists Japanese key in octets and value.'); # XXX
+    is_deeply($sth->{NAME}, [ 'name', Encode::encode('UTF-8', $jpnColumn) ], 'Statement Japanese column name is in octets.'); # XXX
+    is_deeply($sth->{mysql_table}, [ Encode::encode('UTF-8', $jpnTable), Encode::encode('UTF-8', $jpnTable) ], 'Statement Japanese table name is in octets.'); # XXX
+    if ($mysql_enable_utf8 == 0 || $mysql_enable_utf8 == 2) {
+        ok($sth = $dbh->prepare("SELECT * FROM $chiTable_db"), 'Prepare select statement with Chinese values.');
+        ok($sth->execute(), 'Execute select statement with Chinese values.');
+        ok($row = $sth->fetchrow_hashref(), 'Fetch hashref with Chinese values.');
+        is($row->{name}, $chiData1_db, "Chinese value.");
+        ok(!exists $row->{$chiColumn}, 'Not exists Chinese key in internal Perl Unicode.'); # XXX
+        is($row->{Encode::encode('UTF-8', $chiColumn)}, $chiData2_db, 'Exists Chinese key in octets and value.'); # XXX
+        is_deeply($sth->{NAME}, [ 'name', Encode::encode('UTF-8', $chiColumn) ], 'Statement Chinese column name is in octets.'); # XXX
+        is_deeply($sth->{mysql_table}, [ Encode::encode('UTF-8', $chiTable), Encode::encode('UTF-8', $chiTable) ], 'Statement Chinese table name is in octets.'); # XXX
+    }
+
+    $sth->finish();
+    $dbh->disconnect();
+}
+done_testing;