diff --git a/doc/management.txt b/doc/management.txt
index 9a749e4..151c96e 100644
--- a/doc/management.txt
+++ b/doc/management.txt
@@ -961,10 +961,14 @@
 do not insert any column before these ones in order not to break tools which
 use hard-coded column positions.
 
-In brackets after each field name are the types which may have a value for
-that field. The types are L (Listeners), F (Frontends), B (Backends), and
-S (Servers).
+For proxy statistics, after each field name, the types which may have a value
+for that field are specified in brackets. The types are L (Listeners), F
+(Frontends), B (Backends), and S (Servers). There is a fixed set of static
+fields that are always available in the same order. A column containing the
+character '-' delimits the end of the static fields, after which presence or
+order of the fields are not guaranteed.
 
+Here is the list of static fields using the proxy statistics domain:
   0. pxname [LFBS]: proxy name
   1. svname [LFBS]: service name (FRONTEND for frontend, BACKEND for backend,
      any name for server/listener)
@@ -1122,6 +1126,9 @@
  93. ttime_max [..BS]: the maximum observed total session time in ms
  94. eint [LFBS]: cumulative number of internal errors
 
+For all other statistics domains, the presence or the order of the fields are
+not guaranteed. In this case, the header line should always be used to parse
+the CSV data.
 
 9.2) Typed output format
 ------------------------
diff --git a/src/stats.c b/src/stats.c
index 6ab1c06..ed0abc7 100644
--- a/src/stats.c
+++ b/src/stats.c
@@ -324,14 +324,21 @@
  * NOTE: Some tools happen to rely on the field position instead of its name,
  *       so please only append new fields at the end, never in the middle.
  */
-static void stats_dump_csv_header()
+static void stats_dump_csv_header(enum stats_domain domain)
 {
 	int field;
 
 	chunk_appendf(&trash, "# ");
-	for (field = 0; field < ST_F_TOTAL_FIELDS; field++)
+	for (field = 0; field < ST_F_TOTAL_FIELDS; field++) {
 		chunk_appendf(&trash, "%s,", stat_fields[field].name);
 
+		/* print special delimiter on proxy stats to mark end of
+		   static fields */
+		if (domain == STATS_DOMAIN_PROXY && field + 1 == ST_F_TOTAL_FIELDS) {
+			chunk_appendf(&trash, "-,");
+		}
+	}
+
 	chunk_appendf(&trash, "\n");
 }
 
@@ -523,16 +530,25 @@
 /* Dump all fields from <stats> into <out> using CSV format */
 static int stats_dump_fields_csv(struct buffer *out,
                                  const struct field *stats, size_t stats_count,
-                                 unsigned int flags)
+                                 unsigned int flags,
+                                 enum stats_domain domain)
 {
 	int field;
 
-	for (field = 0; field < stats_count; field++) {
+	for (field = 0; field < stats_count; ++field) {
 		if (!stats_emit_raw_data_field(out, &stats[field]))
 			return 0;
 		if (!chunk_strcat(out, ","))
 			return 0;
+
+		/* print special delimiter on proxy stats to mark end of
+		   static fields */
+		if (domain == STATS_DOMAIN_PROXY && field + 1 == ST_F_TOTAL_FIELDS) {
+			if (!chunk_strcat(out, "-,"))
+				return 0;
+		}
 	}
+
 	chunk_strcat(out, "\n");
 	return 1;
 }
@@ -1435,7 +1451,7 @@
 	else if (appctx->ctx.stats.flags & STAT_FMT_JSON)
 		ret = stats_dump_fields_json(&trash, stats, stats_count, appctx->ctx.stats.flags, appctx->ctx.stats.domain);
 	else
-		ret = stats_dump_fields_csv(&trash, stats, stats_count, appctx->ctx.stats.flags);
+		ret = stats_dump_fields_csv(&trash, stats, stats_count, appctx->ctx.stats.flags, appctx->ctx.stats.domain);
 
 	if (ret)
 		appctx->ctx.stats.flags |= STAT_STARTED;
@@ -2837,7 +2853,7 @@
 		else if (appctx->ctx.stats.flags & STAT_FMT_JSON)
 			stats_dump_json_header();
 		else if (!(appctx->ctx.stats.flags & STAT_FMT_TYPED))
-			stats_dump_csv_header();
+			stats_dump_csv_header(appctx->ctx.stats.domain);
 
 		if (!stats_putchk(rep, htx, &trash))
 			goto full;