diff --git a/api/app/models/place.ts b/api/app/models/place.ts
index a03ef860..aad1680d 100644
--- a/api/app/models/place.ts
+++ b/api/app/models/place.ts
@@ -471,7 +471,9 @@ const seedData = async (directory: string) => {
 
   const typeRegex = /^.*\//;
   const parseOsgb = (val: string) => parseInt(val, 10) || "";
+  const stripBom = (val: string) => val.replace(/^\uFEFF/, "");
   const transformExceptions = {
+    code: stripBom,
     northings: parseOsgb,
     eastings: parseOsgb,
     min_eastings: parseOsgb,
diff --git a/test/place.unit.ts b/test/place.unit.ts
index eb225e34..93f103df 100644
--- a/test/place.unit.ts
+++ b/test/place.unit.ts
@@ -58,6 +58,18 @@ describe("Place Model", () => {
         const result = await query(q);
         assert.equal(result.rows[0].count, placesEntriesCount);
       });
+
+      it("strips BOM characters from code field", async () => {
+        // BOM (U+FEFF) can appear at start of CSV files and pollute first field
+        const q = `SELECT code FROM ${Place.relation.relation}`;
+        const result = await query(q);
+        result.rows.forEach((row) => {
+          assert.isFalse(
+            row.code.startsWith("\uFEFF"),
+            `code "${row.code}" should not start with BOM character`
+          );
+        });
+      });
     });
 
     describe("#populateLocation", () => {
diff --git a/test/postcode.regression.unit.ts b/test/postcode.regression.unit.ts
index 4c8c74b3..c11c8a53 100644
--- a/test/postcode.regression.unit.ts
+++ b/test/postcode.regression.unit.ts
@@ -155,7 +155,7 @@ describe("Postcode data regression testing", function () {
         admin_district: "Channel Islands",
         parish: null,
         admin_county: null,
-        admin_ward: null,
+        admin_ward: "Channel Islands (pseudo)",
         ccg: null,
         ccg_code: null,
         nuts: null,