diff --git a/_FiveSql2/src/TSqlAgg.prg b/_FiveSql2/src/TSqlAgg.prg
index 112e110..95f0783 100644
--- a/_FiveSql2/src/TSqlAgg.prg
+++ b/_FiveSql2/src/TSqlAgg.prg
@@ -16,6 +16,7 @@ CLASS TSqlAgg
 
    METHOD New() CONSTRUCTOR
    METHOD GroupBy( aRows, aFN, aCols, aGroupBy, xHaving, aTables, aParams )
+   METHOD FindGroupIdx( xGroupExpr, aCols, aFN )
    METHOD ComputeAgg( xE, aGR, aFN )
    METHOD FindColIdx( xExpr, aFN )
    METHOD FindColIdx2( cN, aFN )
@@ -49,6 +50,7 @@ METHOD GroupBy( aRows, aFN, aCols, aGroupBy, xHaving, aTables, aParams ) CLASS T
    LOCAL i, j, cKey, aGroupRows, aResult := {}
    LOCAL aNewRow
    LOCAL nGCol, cN, nCI, lPass
+   LOCAL aGroupIdx := {}
 
    /* Aggregate on empty set */
    IF Len( aRows ) == 0 .AND. ::HasAgg( aCols )
@@ -63,14 +65,27 @@ METHOD GroupBy( aRows, aFN, aCols, aGroupBy, xHaving, aTables, aParams ) CLASS T
       RETURN { aNewRow }
    ENDIF
 
-   /* Build group buckets */
+   /* Build group buckets.
+    * Pre-resolve the GROUP BY columns to their position in the SELECT
+    * list by matching against the SOURCE expressions in aCols, not the
+    * alias list in aFN. Matching on aFN breaks as soon as the user
+    * writes `SELECT d.name AS foo ... GROUP BY d.name` — the group
+    * column's ND_COL name "D.NAME" wouldn't appear in aFN (which has
+    * "FOO"), FindColIdx would return 0, and every row would end up in
+    * the empty-key bucket collapsing to a single output row.
+    * (Regression found in complex-query bench 2026-04-14.) */
+   FOR j := 1 TO Len( aGroupBy )
+      nGCol := ::FindGroupIdx( aGroupBy[ j ], aCols, aFN )
+      AAdd( aGroupIdx, nGCol )
+   NEXT
+
    IF Len( aGroupBy ) == 0 .AND. ::HasAgg( aCols )
       hGroups[ "__ALL__" ] := aRows
    ELSE
       FOR i := 1 TO Len( aRows )
          cKey := ""
          FOR j := 1 TO Len( aGroupBy )
-            nGCol := ::FindColIdx( aGroupBy[ j ], aFN )
+            nGCol := aGroupIdx[ j ]
             IF nGCol > 0 .AND. nGCol <= Len( aRows[ i ] )
                cKey += SqlValToStr( aRows[ i ][ nGCol ] ) + "|"
             ENDIF
@@ -113,6 +128,45 @@ METHOD GroupBy( aRows, aFN, aCols, aGroupBy, xHaving, aTables, aParams ) CLASS T
 RETURN aResult
 
 
+/* Resolve a GROUP BY expression to its column position in the output row.
+ * Walks the SELECT list's source expressions (aCols[i][1]) rather than
+ * the alias list (aFN[i]). For `SELECT d.name AS foo GROUP BY d.name`,
+ * aFN is {"FOO"} but aCols[1][1] is ND_COL "d.name" — we need to match
+ * the latter, otherwise the group key collapses every row into one
+ * bucket. Falls back to FindColIdx (alias/name lookup) for cases where
+ * the GROUP BY uses a simple identifier that isn't in the SELECT list.
+ */
+METHOD FindGroupIdx( xGroupExpr, aCols, aFN ) CLASS TSqlAgg
+
+   LOCAL i, xSel, cGName, cSName, nDot
+
+   IF xGroupExpr == NIL .OR. xGroupExpr[ 1 ] != ND_COL
+      RETURN ::FindColIdx( xGroupExpr, aFN )
+   ENDIF
+
+   cGName := Upper( xGroupExpr[ 2 ] )
+   nDot := At( ".", cGName )
+   IF nDot > 0
+      cGName := SubStr( cGName, nDot + 1 )
+   ENDIF
+
+   FOR i := 1 TO Len( aCols )
+      xSel := aCols[ i ][ 1 ]
+      IF xSel != NIL .AND. xSel[ 1 ] == ND_COL
+         cSName := Upper( xSel[ 2 ] )
+         IF "." $ cSName
+            cSName := SubStr( cSName, At( ".", cSName ) + 1 )
+         ENDIF
+         IF cSName == cGName
+            RETURN i
+         ENDIF
+      ENDIF
+   NEXT
+
+   /* Last resort: alias-based lookup (handles GROUP BY of unrelated cols) */
+RETURN ::FindColIdx( xGroupExpr, aFN )
+
+
 METHOD FindColIdx( xExpr, aFN ) CLASS TSqlAgg
 
    LOCAL cN, i