diff --git a/checker/cost.go b/checker/cost.go
index 7312d1fe..16e5ecf8 100644
--- a/checker/cost.go
+++ b/checker/cost.go
@@ -340,6 +340,8 @@ func (c *coster) costSelect(e *exprpb.Expr) CostEstimate {
 	sel := e.GetSelectExpr()
 	var sum CostEstimate
 	if sel.GetTestOnly() {
+		sum.Min = 1
+		sum.Max = 1
 		return sum
 	}
 	sum = sum.Add(c.cost(sel.GetOperand()))
diff --git a/checker/cost_test.go b/checker/cost_test.go
index 4e8cb9b4..8948d690 100644
--- a/checker/cost_test.go
+++ b/checker/cost_test.go
@@ -75,7 +75,7 @@ func TestCost(t *testing.T) {
 			name:   "select: field test only",
 			expr:   `has(input.single_int32)`,
 			decls:  []*exprpb.Decl{decls.NewVar("input", decls.NewObjectType("google.expr.proto3.test.TestAllTypes"))},
-			wanted: zeroCost,
+			wanted: CostEstimate{Min: 1, Max: 1},
 		},
 		{
 			name:   "estimated function call",
diff --git a/interpreter/interpretable.go b/interpreter/interpretable.go
index f9724d46..74aef257 100644
--- a/interpreter/interpretable.go
+++ b/interpreter/interpretable.go
@@ -110,11 +110,10 @@ type InterpretableConstructor interface {
 // Core Interpretable implementations used during the program planning phase.
 
 type evalTestOnly struct {
-	id           int64
-	attr         InterpretableAttribute
-	qual         Qualifier
-	field        types.String
-	isProtoField bool
+	id    int64
+	attr  InterpretableAttribute
+	qual  Qualifier
+	field types.String
 }
 
 // ID implements the Interpretable interface method.
diff --git a/interpreter/planner.go b/interpreter/planner.go
index 165a388b..620f9aac 100644
--- a/interpreter/planner.go
+++ b/interpreter/planner.go
@@ -220,19 +220,11 @@ func (p *planner) planSelect(expr *exprpb.Expr) (Interpretable, error) {
 
 	// Return the test only eval expression.
 	if sel.GetTestOnly() {
-		isProtoField := false
-		if opType.GetMessageType() != "" {
-			ft, found := p.provider.FindFieldType(opType.GetMessageType(), sel.GetField())
-			if found && ft.IsSet != nil && ft.GetFrom != nil {
-				isProtoField = true
-			}
-		}
 		return &evalTestOnly{
-			id:           expr.GetId(),
-			field:        types.String(sel.GetField()),
-			attr:         attr,
-			qual:         qual,
-			isProtoField: isProtoField,
+			id:    expr.GetId(),
+			field: types.String(sel.GetField()),
+			attr:  attr,
+			qual:  qual,
 		}, nil
 	}
 
diff --git a/interpreter/runtimecost.go b/interpreter/runtimecost.go
index 33a4ea56..d355733b 100644
--- a/interpreter/runtimecost.go
+++ b/interpreter/runtimecost.go
@@ -38,12 +38,6 @@ type ActualCostEstimator interface {
 func CostObserver(tracker *CostTracker) EvalObserver {
 	observer := func(id int64, programStep any, val ref.Val) {
 		switch t := programStep.(type) {
-		case *evalTestOnly:
-			// Protobuf has() checks do not incur runtime cost due to not using Eval()
-			// in older versions of CEL; this keeps the runtime cost behavior consistent
-			if !t.isProtoField {
-				tracker.cost++
-			}
 		case ConstantQualifier:
 			// TODO: Push identifiers on to the stack before observing constant qualifiers that apply to them
 			// and enable the below pop. Once enabled this can case can be collapsed into the Qualifier case.
@@ -75,6 +69,8 @@ func CostObserver(tracker *CostTracker) EvalObserver {
 			tracker.stack.drop(t.rhs.ID(), t.lhs.ID())
 		case *evalFold:
 			tracker.stack.drop(t.iterRange.ID())
+		case *evalTestOnly:
+			tracker.cost += common.SelectAndIdentCost
 		case Qualifier:
 			tracker.cost++
 		case InterpretableCall:
diff --git a/interpreter/runtimecost_test.go b/interpreter/runtimecost_test.go
index 0bcd39e3..26a19368 100644
--- a/interpreter/runtimecost_test.go
+++ b/interpreter/runtimecost_test.go
@@ -306,7 +306,7 @@ func TestRuntimeCost(t *testing.T) {
 			name:  "select: field test only",
 			expr:  `has(input.single_int32)`,
 			decls: []*exprpb.Decl{decls.NewVar("input", decls.NewObjectType("google.expr.proto3.test.TestAllTypes"))},
-			want:  0,
+			want:  1,
 			in: map[string]any{
 				"input": &proto3pb.TestAllTypes{
 					RepeatedBool: []bool{false},