diff --git a/Makefile b/Makefile
index fa72836..cdea62f 100644
--- a/Makefile
+++ b/Makefile
@@ -26,7 +26,7 @@ else
 endif
 
 EXTENSION = pg_csv
-EXTVERSION = 0.3
+EXTVERSION = 0.4
 
 DATA = $(wildcard sql/*--*.sql)
 
diff --git a/README.md b/README.md
index 06bf374..5f429cb 100644
--- a/README.md
+++ b/README.md
@@ -71,3 +71,19 @@ select csv_agg(x, csv_options(bom := true)) from projects x;
  5,Orphan,
 (1 row)
 ```
+
+### Header
+
+You can omit or include the CSV header.
+
+```psql
+select csv_agg(x, csv_options(header := false)) from projects x;
+      csv_agg
+-------------------
+ 1,Windows 7,1    +
+ 2,Windows 10,1   +
+ 3,IOS,2          +
+ 4,OSX,2          +
+ 5,Orphan,
+(1 row)
+```
diff --git a/sql/pg_csv--0.3--0.4.sql b/sql/pg_csv--0.3--0.4.sql
new file mode 100644
index 0000000..472d56d
--- /dev/null
+++ b/sql/pg_csv--0.3--0.4.sql
@@ -0,0 +1,9 @@
+alter type csv_options add attribute header bool;
+
+create or replace function csv_options(
+  delimiter "char" default NULL,
+  bom       bool default NULL,
+  header    bool default NULL
+) returns csv_options as $$
+  select row(delimiter, bom, header)::csv_options;
+$$ language sql;
diff --git a/sql/pg_csv.sql b/sql/pg_csv.sql
index 9932321..8fed440 100644
--- a/sql/pg_csv.sql
+++ b/sql/pg_csv.sql
@@ -1,10 +1,15 @@
 create type csv_options as (
   delimiter "char"
 , bom       bool
+, header    bool
 );
 
-create or replace function csv_options(delimiter "char" default NULL, bom bool default NULL) returns csv_options as $$
-  select row(delimiter, bom)::csv_options;
+create or replace function csv_options(
+  delimiter "char" default NULL,
+  bom       bool default NULL,
+  header    bool default NULL
+) returns csv_options as $$
+  select row(delimiter, bom, header)::csv_options;
 $$ language sql;
 
 create function csv_agg_transfn(internal, anyelement)
diff --git a/src/pg_csv.c b/src/pg_csv.c
index 21535c2..9570a3f 100644
--- a/src/pg_csv.c
+++ b/src/pg_csv.c
@@ -10,6 +10,7 @@ static const char BOM[3]  = "\xEF\xBB\xBF";
 typedef struct {
   char delim;
   bool with_bom;
+  bool header;
 } CsvOptions;
 
 typedef struct {
@@ -59,14 +60,15 @@ static void parse_csv_options(HeapTupleHeader opts_hdr, CsvOptions *csv_opts) {
   // defaults
   csv_opts->delim    = ',';
   csv_opts->with_bom = false;
+  csv_opts->header   = true;
 
   if (opts_hdr == NULL) return;
 
   TupleDesc desc = lookup_rowtype_tupdesc(HeapTupleHeaderGetTypeId(opts_hdr),
                                           HeapTupleHeaderGetTypMod(opts_hdr));
 
-  Datum values[2];
-  bool  nulls[2];
+  Datum values[3];
+  bool  nulls[3];
 
   heap_deform_tuple(
       &(HeapTupleData){.t_len = HeapTupleHeaderGetDatumLength(opts_hdr), .t_data = opts_hdr}, desc,
@@ -84,6 +86,10 @@ static void parse_csv_options(HeapTupleHeader opts_hdr, CsvOptions *csv_opts) {
     csv_opts->with_bom = DatumGetBool(values[1]);
   }
 
+  if (!nulls[2]) {
+    csv_opts->header = DatumGetBool(values[2]);
+  }
+
   ReleaseTupleDesc(desc);
 }
 
@@ -128,19 +134,21 @@ Datum csv_agg_transfn(PG_FUNCTION_ARGS) {
     if (state->options->with_bom) appendBinaryStringInfo(&state->accum_buf, BOM, sizeof(BOM));
 
     // build header row
-    for (int i = 0; i < tdesc->natts; i++) {
-      Form_pg_attribute att = TupleDescAttr(tdesc, i);
-      if (att->attisdropped) // pg always keeps dropped columns, guard against this
-        continue;
+    if (state->options->header) {
+      for (int i = 0; i < tdesc->natts; i++) {
+        Form_pg_attribute att = TupleDescAttr(tdesc, i);
+        if (att->attisdropped) // pg always keeps dropped columns, guard against this
+          continue;
 
-      if (i > 0) // only append delimiter after the first value
-        appendStringInfoChar(&state->accum_buf, state->options->delim);
+        if (i > 0) // only append delimiter after the first value
+          appendStringInfoChar(&state->accum_buf, state->options->delim);
 
-      char *cstr = NameStr(att->attname);
-      csv_append_field(&state->accum_buf, cstr, strlen(cstr), state->options->delim);
-    }
+        char *cstr = NameStr(att->attname);
+        csv_append_field(&state->accum_buf, cstr, strlen(cstr), state->options->delim);
+      }
 
-    appendStringInfoChar(&state->accum_buf, NEWLINE);
+      appendStringInfoChar(&state->accum_buf, NEWLINE);
+    }
 
     state->tupdesc     = tdesc;
     state->header_done = true;
diff --git a/test/expected/bom.out b/test/expected/bom.out
index f339569..0b2ea3d 100644
--- a/test/expected/bom.out
+++ b/test/expected/bom.out
@@ -36,3 +36,7 @@ FROM   projects x;
 7;"has 
 CR";8
 8;"has 
  CRLF""";8
+\echo
+
+\pset format aligned
+\pset tuples_only off
diff --git a/test/expected/header.out b/test/expected/header.out
new file mode 100644
index 0000000..6ca8e79
--- /dev/null
+++ b/test/expected/header.out
@@ -0,0 +1,78 @@
+-- header
+SELECT csv_agg(x, csv_options(header:=true)) AS body
+FROM   projects x;
+             body              
+-------------------------------
+ id,name,client_id            +
+ 1,Windows 7,1                +
+ 2,"has,comma",1              +
+ ,,                           +
+ 4,OSX,2                      +
+ ,"has""quote",               +
+ 5,"has,comma and ""quote""",7+
+ 6,"has                       +
+  LF",7                       +
+ 7,"has \r CR",8              +
+ 8,"has \r                    +
+  CRLF""",8
+(1 row)
+
+-- no header
+SELECT csv_agg(x, csv_options(header:=false)) AS body
+FROM   projects x;
+             body              
+-------------------------------
+ 1,Windows 7,1                +
+ 2,"has,comma",1              +
+ ,,                           +
+ 4,OSX,2                      +
+ ,"has""quote",               +
+ 5,"has,comma and ""quote""",7+
+ 6,"has                       +
+  LF",7                       +
+ 7,"has \r CR",8              +
+ 8,"has \r                    +
+  CRLF""",8
+(1 row)
+
+-- no header with delimiter
+SELECT csv_agg(x, csv_options(delimiter:='|', header:=false)) AS body
+FROM   projects x;
+             body              
+-------------------------------
+ 1|Windows 7|1                +
+ 2|has,comma|1                +
+ ||                           +
+ 4|OSX|2                      +
+ |"has""quote"|               +
+ 5|"has,comma and ""quote"""|7+
+ 6|"has                       +
+  LF"|7                       +
+ 7|"has \r CR"|8              +
+ 8|"has \r                    +
+  CRLF"""|8
+(1 row)
+
+-- see bom.sql for an explanation of these settings
+\pset format unaligned
+\pset tuples_only on
+\echo
+
+-- no header with delimiter and BOM
+SELECT csv_agg(x, csv_options(delimiter:='|', header:=false, bom := true)) AS body
+FROM   projects x;
+1|Windows 7|1
+2|has,comma|1
+||
+4|OSX|2
+|"has""quote"|
+5|"has,comma and ""quote"""|7
+6|"has 
+ LF"|7
+7|"has 
 CR"|8
+8|"has 
+ CRLF"""|8
+\echo
+
+\pset format aligned
+\pset tuples_only off
diff --git a/test/sql/bom.sql b/test/sql/bom.sql
index 60f086e..a9f98d6 100644
--- a/test/sql/bom.sql
+++ b/test/sql/bom.sql
@@ -12,3 +12,7 @@ FROM   projects x;
 -- include BOM with custom delimiter
 SELECT csv_agg(x, csv_options(delimiter := ';', bom := true)) AS body
 FROM   projects x;
+\echo
+
+\pset format aligned
+\pset tuples_only off
diff --git a/test/sql/header.sql b/test/sql/header.sql
new file mode 100644
index 0000000..3f73573
--- /dev/null
+++ b/test/sql/header.sql
@@ -0,0 +1,24 @@
+-- header
+SELECT csv_agg(x, csv_options(header:=true)) AS body
+FROM   projects x;
+
+-- no header
+SELECT csv_agg(x, csv_options(header:=false)) AS body
+FROM   projects x;
+
+-- no header with delimiter
+SELECT csv_agg(x, csv_options(delimiter:='|', header:=false)) AS body
+FROM   projects x;
+
+-- see bom.sql for an explanation of these settings
+\pset format unaligned
+\pset tuples_only on
+\echo
+
+-- no header with delimiter and BOM
+SELECT csv_agg(x, csv_options(delimiter:='|', header:=false, bom := true)) AS body
+FROM   projects x;
+\echo
+
+\pset format aligned
+\pset tuples_only off