If you what you want (as hinted by Re^2: Angle brackets and regex problem) is to match the longest match to /\d+[a-z]*/ which is not surrounded by > and <, you may say:

my $test = "blah <span class='small'>13a</span> blah blah";
if ( $test =~ /(.)(\d+[a-z]*)(.)/ && $1 ne '>' && $2 ne '<') {
    print "No tag found: $2\n";
} else {
    print "No match"
}
[download]

and it is going to say "No match". While using my $test = "blah 13a blah blah"; would result

No tag found: 13a
[download]

But if you want something more complex than this — like finding the content of XML tags which just contain text which matches /\d+[a-z]*/, listen to what merlyn said and use a HTML parser.