diff --git a/packages/scenario/test/calculate.test.ts b/packages/scenario/test/calculate.test.ts
new file mode 100644
index 00000000..68239686
--- /dev/null
+++ b/packages/scenario/test/calculate.test.ts
@@ -0,0 +1,52 @@
+import {
+	bind,
+	body,
+	head,
+	html,
+	input,
+	mainInstance,
+	model,
+	t,
+	title,
+} from '@odk-web-forms/common/test/fixtures/xform-dsl/index.ts';
+import { describe, expect, it } from 'vitest';
+import { intAnswer } from '../src/answer/ExpectedIntAnswer.ts';
+import { Scenario } from '../src/jr/Scenario.ts';
+
+describe('TriggerableDagTest.java', () => {
+	/**
+	 * **PORTING NOTES**
+	 *
+	 * Rephrase? While DAG ordering is certainly under test, and while it's an
+	 * explicit spec concern, there's also more specific computation logic under
+	 * test which is worth describing as the functionality under test.
+	 */
+	it('[recomputes `calculate` expressions when their dependencies are updated] order of the DAG is ensured', async () => {
+		const scenario = await Scenario.init(
+			'Some form',
+			html(
+				head(
+					title('Some form'),
+					model(
+						mainInstance(t('data id="some-form"', t('a', '2'), t('b'), t('c'))),
+						bind('/data/a').type('int'),
+						bind('/data/b').type('int').calculate('/data/a * 3'),
+						bind('/data/c').type('int').calculate('(/data/a + /data/b) * 5')
+					)
+				),
+				body(input('/data/a'))
+			)
+		);
+
+		expect(scenario.answerOf('/data/a')).toEqualAnswer(intAnswer(2));
+		expect(scenario.answerOf('/data/b')).toEqualAnswer(intAnswer(6));
+		expect(scenario.answerOf('/data/c')).toEqualAnswer(intAnswer(40));
+
+		scenario.answer('/data/a', 3);
+
+		expect(scenario.answerOf('/data/a')).toEqualAnswer(intAnswer(3));
+		expect(scenario.answerOf('/data/b')).toEqualAnswer(intAnswer(9));
+		// Verify that c gets computed using the updated value of b.
+		expect(scenario.answerOf('/data/c')).toEqualAnswer(intAnswer(60));
+	});
+});