diff --git a/spoonbill/flatten.py b/spoonbill/flatten.py
index a884a45..9646ad9 100644
--- a/spoonbill/flatten.py
+++ b/spoonbill/flatten.py
@@ -29,6 +29,7 @@ class TableFlattenConfig:
     only: List[str] = field(default_factory=list)
     repeat: List[str] = field(default_factory=list)
     unnest: List[str] = field(default_factory=list)
+    only: List[str] = field(default_factory=list)
     name: str = ""
 
 
@@ -58,7 +59,7 @@ class Flattener:
     In order to export data correctly Flattener requires previously analyzed tables data.
     During the process flattener could add columns not based on schema analysis, such as
     `itemsCount`.
-    In every generated row, depending on table type, flattener will always few add augenerated columns.
+    In every generated row, depending on table type, flattener will always few add autogenerated columns.
     For root table:
     * rowID
     * id
@@ -100,12 +101,15 @@ def _init_table_cache(self, tables, table):
 
     def _init_options(self, tables):
         for table in tables.values():
+
             name = table.name
             count = self.options.count
             options = self.options.selection[name]
             unnest = options.unnest
             split = options.split
             repeat = options.repeat
+            only = options.only
+
             if count:
                 for array in table.arrays:
                     parts = array.split("/")
@@ -142,6 +146,19 @@ def _init_options(self, tables):
                         child_table = self.tables.get(c_name)
                         child_table.columns[col_id] = col
                         child_table.titles[col_id] = title
+            if only:
+                if split:
+                    table.columns = {c_id: c for c_id, c in table.columns.items() if c_id in only}
+                else:
+                    table.combined_columns = {c_id: c for c_id, c in table.combined_columns.items() if c_id in only}
+
+    def _only(self, table, only, split):
+        table.types = {c_id: c for c_id, c in table.types.items() if c_id in only}
+
+        if split:
+            table.columns = {c_id: c for c_id, c in table.columns.items() if c_id in only}
+            return
+        table.combined_columns = {c_id: c for c_id, c in table.combined_columns.items() if c_id in only}
 
     def _init(self):
         # init cache and filter only selected tables
@@ -149,8 +166,11 @@ def _init(self):
         for name, table in self.tables.items():
             if name not in self.options.selection:
                 continue
+            options = self.options.selection[name]
+            split = options.split
+            if options.only:
+                self._only(table, options.only, split)
             self._init_table_cache(tables, table)
-            split = self.options.selection[name].split
             if split:
                 for c_name in table.child_tables:
                     if c_name in self.options.exclude:
diff --git a/tests/test_flatten.py b/tests/test_flatten.py
index e929778..4cc993f 100644
--- a/tests/test_flatten.py
+++ b/tests/test_flatten.py
@@ -120,11 +120,20 @@ def test_flatten_with_exclude(spec_analyzed, releases):
 
 
 def test_flatten_with_only(spec_analyzed, releases):
-    options = FlattenOptions(**{"selection": {"tenders": {"split": True}}, "exclude": "tender_items"})
+    options = FlattenOptions(**{"selection": {"tenders": {"split": True, "only": ["/tender/id"]}}})
     flattener = Flattener(options, spec_analyzed.tables)
     all_rows = defaultdict(list)
     for count, flat in flattener.flatten(releases):
         for name, rows in flat.items():
             all_rows[name].extend(rows)
+    for row in all_rows["tenders"]:
+        assert row == ["/tender/id"]
 
-    assert "tender_items" not in all_rows
+    options = FlattenOptions(**{"selection": {"tenders": {"split": False, "only": ["/tender/id"]}}})
+    flattener = Flattener(options, spec_analyzed.tables)
+    all_rows = defaultdict(list)
+    for count, flat in flattener.flatten(releases):
+        for name, rows in flat.items():
+            all_rows[name].extend(rows)
+    for row in all_rows["tenders"]:
+        assert row == ["/tender/id"]