diff --git a/README.md b/README.md
index 31d20f6..d325dd6 100644
--- a/README.md
+++ b/README.md
@@ -263,12 +263,12 @@ The default name format is: `"{Callsign} {Nearest City} {Landmark}"`
 
 # Bonus: Trim the Contact List
 
-Download the usersDB.bin with Farnsworth editcp.
+For MD-UV380 only.
 
     python -m dzcb.contact_trim \
-        < "~/.cache/codeplug/Codeplug Editor/usersDB.bin" \
-        > "~/.cache/codeplug/Codeplug Editor/usersDB-trimmed.bin"
+        "~/.cache/codeplug/Codeplug Editor/usersDB.bin"
 
+Write the database with editcp (do NOT check Download New)
 
 # Basic Usage
 
diff --git a/src/dzcb/contacts_trim.py b/src/dzcb/contacts_trim.py
index bc3fb1e..f9afd09 100644
--- a/src/dzcb/contacts_trim.py
+++ b/src/dzcb/contacts_trim.py
@@ -1,57 +1,68 @@
 """
 dzcb.contacts_trim - remove contacts to get under radio limits
 """
+import argparse
+import csv
+import json
+from pathlib import Path
+import time
 import sys
 
-remove_suffixes = [
-    ",",
-    ",GR",
-    ",CY",
-    ",CN",
-    ",BE",
-    ",FR",
-    ",ES",
-    ",IT",
-    ",RU",
-    ",PL",
-    ",DE",
-    ",PT",
-    ",TR",
-    ",SI",
-    ",JP",
-    ",Korea Republic of",
-    ",PH",
-    ",MY",
-    ",TH",
-    ",AR",
-    ",BR",
-    ",CL",
-    ",CO",
-    ",VE",
-    ",UY",
-    ",SE",
-    ",CH",
-    ",CZ",
-    ",SK",
-    "BA",
-    "HR",
-    "AT",
-]
+import requests
+
+from dzcb import appdir
+
+RADIO_ID_USERS_JSON = "https://database.radioid.net/static/users.json"
+RADIO_ID_USERS_MAX_AGE = 3600 * 12.1
+
 
 total = 0
 
 
-def check_suffix(line):
-    for s in remove_suffixes:
-        if line.strip().endswith(s):
-            return None
-    return line
+def cached_json(url, max_age=RADIO_ID_USERS_MAX_AGE):
+    cachedir = Path(appdir.user_cache_dir)
+    filepath = cachedir / "usersdb.json"
+    if not filepath.exists() or filepath.stat().st_mtime < time.time() - max_age:
+        # cache is expired, need to refetch
+        cachedir.mkdir(parents=True, exist_ok=True)
+        resp = requests.get(url)
+        filepath.write_bytes(resp.content)
+    return filepath
+
+
+def group_users_by(field_name):
+    groups = {}
+    with open(cached_json(RADIO_ID_USERS_JSON)) as f:
+        db = json.load(f)
+        for user in db["users"]:
+            groups.setdefault(user[field_name].lower(), []).append(user)
+    return groups
+
+
+def flatten_groups(groups):
+    users = []
+    for ulist in groups.values():
+        users.extend(ulist)
+    return users
+
 
+def users_to_md_uv380_csv(users, output):
+    fields = ["radio_id", "callsign", "fname", "city", "state", "remarks", "country"]
+    with open(output, "w", newline="") as out:
+        csvw = csv.DictWriter(out, fieldnames=fields, extrasaction="ignore")
+        for u in users:
+            u["fname"] = u["fname"].partition(" ")[0].capitalize()
+            u["city"] = u["city"].capitalize()
+            u["state"] = u["state"].capitalize()
+            u["remarks"] = ""
+            csvw.writerow(u)
 
-for line in sys.stdin:
-    outline = check_suffix(line)
-    if outline:
-        total += 1
-        sys.stdout.write(outline)
 
-sys.stderr.write(f"Wrote {total} records")
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("output_file")
+    args = parser.parse_args()
+    users_by_country = group_users_by("country")
+    filtered = users_by_country["united states"] + users_by_country["canada"]
+    users_to_md_uv380_csv(filtered, args.output_file)
+    sys.stderr.write(f"Wrote {len(filtered)} records")