diff --git a/openerp/import_xml.rng b/openerp/import_xml.rng
index d301ae47ed3..97165f24203 100644
--- a/openerp/import_xml.rng
+++ b/openerp/import_xml.rng
@@ -114,6 +114,10 @@
             <rng:optional> <rng:attribute name="attachment" /> </rng:optional>
             <rng:optional> <rng:attribute name="attachment_use" /> </rng:optional>
             <rng:optional> <rng:attribute name="groups"/> </rng:optional>
+            <!-- `Usage` may help identify the best report for a certain task,
+                 for example usage="default" for reports that are attached by
+                 default in EDI exports -->
+            <rng:optional><rng:attribute name="usage"/></rng:optional>
             <rng:empty />
         </rng:element>
     </rng:define>
diff --git a/openerp/tools/convert.py b/openerp/tools/convert.py
index e0b7cdeeb5f..7d6a982986b 100644
--- a/openerp/tools/convert.py
+++ b/openerp/tools/convert.py
@@ -314,7 +314,8 @@ form: module.record_id""" % (xml_id,)
         for dest,f in (('name','string'),('model','model'),('report_name','name')):
             res[dest] = rec.get(f,'').encode('utf8')
             assert res[dest], "Attribute %s of report is empty !" % (f,)
-        for field,dest in (('rml','report_rml'),('file','report_rml'),('xml','report_xml'),('xsl','report_xsl'),('attachment','attachment'),('attachment_use','attachment_use')):
+        for field,dest in (('rml','report_rml'),('file','report_rml'),('xml','report_xml'),('xsl','report_xsl'),
+                           ('attachment','attachment'),('attachment_use','attachment_use'), ('usage','usage')):
             if rec.get(field):
                 res[dest] = rec.get(field).encode('utf8')
         if rec.get('auto'):