diff --git a/env.example b/env.example
index 4eaaa1b..abe7fc9 100644
--- a/env.example
+++ b/env.example
@@ -3,7 +3,7 @@
 # ===========================================
 
 # Site Configuration
-NEXT_PUBLIC_SITE_URL=https://gnars.com
+NEXT_PUBLIC_SITE_URL=https://www.gnars.com
 NEXT_PUBLIC_BASE_URL=http://localhost:3000
 
 # ===========================================
diff --git a/src/app/robots.ts b/src/app/robots.ts
index 74a109d..2cec847 100644
--- a/src/app/robots.ts
+++ b/src/app/robots.ts
@@ -1,6 +1,9 @@
 import type { MetadataRoute } from "next";
 
-const SITE_URL = process.env.NEXT_PUBLIC_SITE_URL || "https://gnars.com";
+// Canonical host is www.gnars.com (matches metadataBase in [locale]/layout.tsx
+// and BASE_URL in lib/config.ts). Keeping this in sync is what lets Google index
+// the site — a non-www default here made every sitemap URL a redirect to www.
+const SITE_URL = process.env.NEXT_PUBLIC_SITE_URL || "https://www.gnars.com";
 
 export default function robots(): MetadataRoute.Robots {
   const normalized = SITE_URL.replace(/\/+$/, "");
diff --git a/src/app/sitemap.xml/route.ts b/src/app/sitemap.xml/route.ts
index 5547735..f74bc1f 100644
--- a/src/app/sitemap.xml/route.ts
+++ b/src/app/sitemap.xml/route.ts
@@ -30,7 +30,10 @@ const PROPOSAL_PAGE_SIZE = 200;
 const MAX_COIN_PAGES = 50;
 const COIN_PAGE_SIZE = 200;
 
-const SITE_URL = process.env.NEXT_PUBLIC_SITE_URL || "https://gnars.com";
+// Canonical host is www.gnars.com (matches metadataBase in [locale]/layout.tsx
+// and BASE_URL in lib/config.ts). A non-www default here emitted 2,564 URLs that
+// all 307-redirected to www, so Google dropped them — keep this on www.
+const SITE_URL = process.env.NEXT_PUBLIC_SITE_URL || "https://www.gnars.com";
 const BASE = SITE_URL.replace(/\/+$/, "");
 
 const toUrl = (path: string) => new URL(path, `${BASE}/`).toString();