diff --git a/docs/html/md__schemas.html b/docs/html/md__schemas.html
index fe7f56357..de7969cb4 100644
--- a/docs/html/md__schemas.html
+++ b/docs/html/md__schemas.html
@@ -166,6 +166,7 @@ root_type Monster;
 <li>It accepts field names with and without quotes, like many JSON parsers already do. It outputs them without quotes as well, though can be made to output them using the <code>strict_json</code> flag.</li>
 <li>If a field has an enum type, the parser will recognize symbolic enum values (with or without quotes) instead of numbers, e.g. <code>field: EnumVal</code>. If a field is of integral type, you can still use symbolic names, but values need to be prefixed with their type and need to be quoted, e.g. <code>field: "Enum.EnumVal"</code>. For enums representing flags, you may place multiple inside a string separated by spaces to OR them, e.g. <code>field: "EnumVal1 EnumVal2"</code> or <code>field: "Enum.EnumVal1 Enum.EnumVal2"</code>.</li>
 <li>Similarly, for unions, these need to specified with two fields much like you do when serializing from code. E.g. for a field <code>foo</code>, you must add a field <code>foo_type: FooOne</code> right before the <code>foo</code> field, where <code>FooOne</code> would be the table out of the union you want to use.</li>
+<li>A field that has the value <code>null</code> (e.g. <code>field: null</code>) is intended to have the default value for that field (thus has the same effect as if that field wasn't specified at all).</li>
 </ul>
 <p>When parsing JSON, it recognizes the following escape codes in strings:</p>
 <ul>
diff --git a/docs/source/Schemas.md b/docs/source/Schemas.md
index 8d4348cbf..36bebf17d 100755
--- a/docs/source/Schemas.md
+++ b/docs/source/Schemas.md
@@ -313,6 +313,9 @@ JSON:
     you do when serializing from code. E.g. for a field `foo`, you must
     add a field `foo_type: FooOne` right before the `foo` field, where
     `FooOne` would be the table out of the union you want to use.
+-   A field that has the value `null` (e.g. `field: null`) is intended to
+    have the default value for that field (thus has the same effect as if
+    that field wasn't specified at all).
 
 When parsing JSON, it recognizes the following escape codes in strings:
 
diff --git a/src/idl_parser.cpp b/src/idl_parser.cpp
index e4013e8d4..2312c3ac0 100644
--- a/src/idl_parser.cpp
+++ b/src/idl_parser.cpp
@@ -151,7 +151,8 @@ std::string Namespace::GetFullyQualifiedName(const std::string &name,
   TD(FileIdentifier, 267, "file_identifier") \
   TD(FileExtension, 268, "file_extension") \
   TD(Include, 269, "include") \
-  TD(Attribute, 270, "attribute")
+  TD(Attribute, 270, "attribute") \
+  TD(Null, 271, "null")
 #ifdef __GNUC__
 __extension__  // Stop GCC complaining about trailing comma with -Wpendantic.
 #endif
@@ -343,6 +344,10 @@ CheckedError Parser::Next() {
             token_ = kTokenFileExtension;
             return NoError();
           }
+          if (attribute_ == "null") {
+            token_ = kTokenNull;
+            return NoError();
+          }
           // If not, it is a user-defined identifier:
           token_ = kTokenIdentifier;
           return NoError();
@@ -688,19 +693,23 @@ CheckedError Parser::ParseTable(const StructDef &struct_def, std::string *value,
       }
     } else {
       EXPECT(':');
-      Value val = field->value;
-      ECHECK(ParseAnyValue(val, field, fieldn));
-      size_t i = field_stack_.size();
-      // Hardcoded insertion-sort with error-check.
-      // If fields are specified in order, then this loop exits immediately.
-      for (; i > field_stack_.size() - fieldn; i--) {
-        auto existing_field = field_stack_[i - 1].second;
-        if (existing_field == field)
-          return Error("field set more than once: " + field->name);
-        if (existing_field->value.offset < field->value.offset) break;
+      if (Is(kTokenNull)) {
+        NEXT(); // Ignore this field.
+      } else {
+        Value val = field->value;
+        ECHECK(ParseAnyValue(val, field, fieldn));
+        size_t i = field_stack_.size();
+        // Hardcoded insertion-sort with error-check.
+        // If fields are specified in order, then this loop exits immediately.
+        for (; i > field_stack_.size() - fieldn; i--) {
+          auto existing_field = field_stack_[i - 1].second;
+          if (existing_field == field)
+            return Error("field set more than once: " + field->name);
+          if (existing_field->value.offset < field->value.offset) break;
+        }
+        field_stack_.insert(field_stack_.begin() + i, std::make_pair(val, field));
+        fieldn++;
       }
-      field_stack_.insert(field_stack_.begin() + i, std::make_pair(val, field));
-      fieldn++;
     }
     if (Is('}')) { NEXT(); break; }
     EXPECT(',');
@@ -1542,10 +1551,10 @@ CheckedError Parser::SkipJsonObject() {
 
 CheckedError Parser::SkipJsonArray() {
   EXPECT('[');
-  
+
   for (;;) {
     if (Is(']')) break;
-        
+
     ECHECK(SkipAnyJsonValue());
 
     if (Is(']')) break;
diff --git a/tests/monsterdata_test.json b/tests/monsterdata_test.json
index 89278755e..12c02d21c 100755
--- a/tests/monsterdata_test.json
+++ b/tests/monsterdata_test.json
@@ -21,7 +21,8 @@
   ],
   test_type: Monster,
   test: {
-    name: "Fred"
+    name: "Fred",
+    pos: null
   },
   test4: [
     {