diff --git a/Documentation/git-fast-import.txt b/Documentation/git-fast-import.txt
index 30ee98d17f..a92ae6d1a3 100644
--- a/Documentation/git-fast-import.txt
+++ b/Documentation/git-fast-import.txt
@@ -176,6 +176,15 @@ results, such as branch names or file names with leading or trailing
 spaces in their name, or early termination of fast-import when it encounters
 unexpected input.
 
+Stream Comments
+~~~~~~~~~~~~~~~
+To aid in debugging frontends fast-import ignores any line that
+begins with `#` (ASCII pound/hash) up to and including the line
+ending `LF`.  A comment line may contain any sequence of bytes
+that does not contain an LF and therefore may be used to include
+any detailed debugging information that might be specific to the
+frontend and useful when inspecting a fast-import data stream.
+
 Date Formats
 ~~~~~~~~~~~~
 The following date formats are supported.  A frontend should select
@@ -689,6 +698,11 @@ intended for production-quality conversions should always use the
 exact byte count format, as it is more robust and performs better.
 The delimited format is intended primarily for testing fast-import.
 
+Comment lines appearing within the `<raw>` part of `data` commands
+are always taken to be part of the body of the data and are therefore
+never ignored by fast-import.  This makes it safe to import any
+file/message content whose lines might start with `#`.
+
 Exact byte count format::
 	The frontend must specify the number of bytes of data.
 +
diff --git a/fast-import.c b/fast-import.c
index d7fa2b7baa..98ebe4770d 100644
--- a/fast-import.c
+++ b/fast-import.c
@@ -122,6 +122,17 @@ Format of STDIN stream:
   email ::= # valid GIT author/committer email;
   ts    ::= # time since the epoch in seconds, ascii base10 notation;
   tz    ::= # GIT style timezone;
+
+     # note: comments may appear anywhere in the input, except
+     # within a data command.  Any form of the data command
+     # always escapes the related input from comment processing.
+     #
+     # In case it is not clear, the '#' that starts the comment
+     # must be the first character on that the line (an lf have
+     # preceeded it).
+     #
+  comment ::= '#' not_lf* lf;
+  not_lf  ::= # Any byte that is not ASCII newline (LF);
 */
 
 #include "builtin.h"
@@ -1454,7 +1465,9 @@ static void dump_marks(void)
 
 static void read_next_command(void)
 {
-	read_line(&command_buf, stdin, '\n');
+	do {
+		read_line(&command_buf, stdin, '\n');
+	} while (!command_buf.eof && command_buf.buf[0] == '#');
 }
 
 static void cmd_mark(void)
@@ -1481,7 +1494,7 @@ static void *cmd_data (size_t *size)
 		length = 0;
 		buffer = xmalloc(sz);
 		for (;;) {
-			read_next_command();
+			read_line(&command_buf, stdin, '\n');
 			if (command_buf.eof)
 				die("EOF in data (terminator '%s' not found)", term);
 			if (term_len == command_buf.len
diff --git a/t/t9300-fast-import.sh b/t/t9300-fast-import.sh
index dac6135b22..1f6426a49e 100755
--- a/t/t9300-fast-import.sh
+++ b/t/t9300-fast-import.sh
@@ -778,4 +778,44 @@ test_expect_success \
 	'git-fast-import <input &&
 	 test `git-rev-parse N2^{tree}` = `git-rev-parse N3^{tree}`'
 
+###
+### series O
+###
+
+cat >input <<INPUT_END
+#we will
+commit refs/heads/O1
+# -- ignore all of this text
+committer $GIT_COMMITTER_NAME <$GIT_COMMITTER_EMAIL> $GIT_COMMITTER_DATE
+# $GIT_COMMITTER_NAME has inserted here for his benefit.
+data <<COMMIT
+dirty directory copy
+COMMIT
+
+# don't forget the import blank line!
+#
+# yes, we started from our usual base of branch^0.
+# i like branch^0.
+from refs/heads/branch^0
+# and we need to reuse file2/file5 from N3 above.
+M 644 inline file2/file5
+# otherwise the tree will be different
+data <<EOF
+$file5_data
+EOF
+
+# don't forget to copy file2 to file3
+C file2 file3
+#
+# or to delete file5 from file2.
+D file2/file5
+# are we done yet?
+
+INPUT_END
+
+test_expect_success \
+	'O: comments are all skipped' \
+	'git-fast-import <input &&
+	 test `git-rev-parse N3` = `git-rev-parse O1`'
+
 test_done