diff --git a/python/pyspark/sql/context.py b/python/pyspark/sql/context.py
index 79453658a1..924bb6433d 100644
--- a/python/pyspark/sql/context.py
+++ b/python/pyspark/sql/context.py
@@ -318,13 +318,7 @@ class SQLContext(object):
                     struct.names[i] = name
             schema = struct
 
-        elif isinstance(schema, StructType):
-            # take the first few rows to verify schema
-            rows = rdd.take(10)
-            for row in rows:
-                _verify_type(row, schema)
-
-        else:
+        elif not isinstance(schema, StructType):
             raise TypeError("schema should be StructType or list or None, but got: %s" % schema)
 
         # convert python objects to sql data