diff --git a/packages/transition-backend/file/__tests__/definitions.test.ts b/packages/transition-backend/file/__tests__/definitions.test.ts
new file mode 100644
index 00000000..2ef162c9
--- /dev/null
+++ b/packages/transition-backend/file/__tests__/definitions.test.ts
@@ -0,0 +1,109 @@
+import fs from 'fs';
+
+// Language configuration - TODO: this should be moved to a separate config file
+const languages = [
+    'en', // first language is used as the reference for the consistency tests below.
+    'fr'
+];
+
+/*
+    this finds all the labels in the file
+    the labels are used to connect to help popups in the transition UI
+    - {([^}]+)} is searching for content between curly braces {} with specific components:
+    - [^}] means "match any character that is NOT a closing curly brace"
+    - } - Matches a closing curly brace literally
+*/
+const labelRegex = /\\label{([^}]+)}/g;
+
+describe('LaTeX Label Consistency Tests', () => {
+    // Helper function to extract labels from LaTeX content
+    function extractLabels(content) {
+
+        const labels = new Set();
+        let match;
+
+        while ((match = labelRegex.exec(content)) !== null) {
+            labels.add(match[1]);
+        }
+
+        return labels;
+    }
+
+    // Store file contents and labels
+    const fileContents = {};
+    const labelSets = {};
+    const referenceLanguage = languages[0];
+
+    beforeAll(() => {
+        // Load all language files
+        languages.forEach((lang) => {
+            fileContents[lang] = fs.readFileSync(`file/definitions/definitions-${lang}.tex`, 'utf8');
+            labelSets[lang] = extractLabels(fileContents[lang]);
+        });
+    });
+
+    // Test each non-reference language against reference language
+    describe.each(languages.slice(1))('Testing $code against reference language ${referenceLanguage}', (lang) => {
+        test(`all reference language ${referenceLanguage} labels exist in ${lang} file`, () => {
+            const missingLabels = [...labelSets[referenceLanguage]].filter(
+                (label) => !labelSets[lang].has(label)
+            );
+
+            if (missingLabels.length > 0) {
+                throw new Error(
+                    `Missing labels in ${lang}:\n${missingLabels.join('\n')}`
+                );
+            }
+        });
+
+        test(`all ${lang} labels exist in reference language ${referenceLanguage} file`, () => {
+            const extraLabels = [...labelSets[lang]].filter(
+                (label) => !labelSets[referenceLanguage].has(label)
+            );
+
+            if (extraLabels.length > 0) {
+                throw new Error(
+                    `Extra labels in ${lang} not present in reference language ${referenceLanguage}:\n${extraLabels.join('\n')}`
+                );
+            }
+        });
+
+        test(`${lang} has same number of labels as reference language ${referenceLanguage}`, () => {
+            expect(labelSets[lang].size).toBe(labelSets[referenceLanguage].size);
+        });
+
+        test(`labels appear in same order as reference language ${referenceLanguage} in ${lang} file`, () => {
+            const langLabels = [...labelSets[lang]];
+            const referenceLabels = [...labelSets[referenceLanguage]];
+            expect(langLabels).toEqual(referenceLabels);
+        });
+    });
+
+    // Test each language for duplicates
+    test.each(languages)('no duplicate labels in $code file', (lang) => {
+        const duplicates = new Set();
+        const seen = new Set();
+        /*
+            this finds all the labels in the file
+            the labels are used to connect to help popups in the transition UI
+            - {([^}]+)} is searching for content between curly braces {} with specific components:
+            - [^}] means "match any character that is NOT a closing curly brace"
+            - } - Matches a closing curly brace literally
+        */
+        let match;
+
+        while ((match = labelRegex.exec(fileContents[lang])) !== null) {
+            const label = match[1];
+            if (seen.has(label)) {
+                duplicates.add(label);
+            }
+            seen.add(label);
+        }
+
+        if (duplicates.size > 0) {
+            throw new Error(
+                `Duplicate labels found in ${lang}:\n${[...duplicates].join('\n')}`
+            );
+        }
+    });
+});
diff --git a/packages/transition-backend/file/definitions/README.md b/packages/transition-backend/file/definitions/README.md
index 269cbcf0..c1a210aa 100644
--- a/packages/transition-backend/file/definitions/README.md
+++ b/packages/transition-backend/file/definitions/README.md
@@ -9,4 +9,6 @@ When adding or modifying information on these files, they must follow the alread
 \makecell[r]{Title} & \[Symbol\] & \[Unit\] & \[Expression\] & Description \\
 \hline
 ```
+After any edit, make sure labels are consistent in both files. For this, run `yarn workspace transition-backend test definitions`.
+
 Each row is starts with a label that allows for it to be located by the tooltip component and ends with `\hline`. Each cell is separated with `&`, and starts and ends with a space. The unit and expression can be empty, in which case they must be represented by a single hyphen (`-`). The symbol, unit, and expression are mathematical expressions, and must be start `\[` and end with `\]`, unless they are empty. Finally, any change must be translated and added to both the French and English file.
\ No newline at end of file