Lately I've been researching bitcoin cold storage and have become increasingly more interested in Casascius Coins. Unfortunately Mike Caldwell had to suspend selling the coins preloaded w/ btc so now you have to buy them blank and load them w/ the btc yourself.

While exploring the site I found a full list of all Casascius coins ever minted. This was so interesting that of course I started picking out addresses at random and looking them up on the blockchain to check their balance.

I noticed that many of the addresses that I checked either had 0 transactions/balance, which leads me to believe that at least some and perhaps many of the coins in the list never actually had any btc loaded onto them, or had already been 'spent.'

This got me wondering exactly how many of the listed coins had not been spent. So I decided to write a short script to query the block chain and figure it out (yay transparency!).

First I started with's Address Balance API because it accepts multiple input values. After experimenting it appeared to be able to return the balance of up 20 wallets at a time without falling over.

So I wrote a ruby script which pulls the values out of the fulllist.txt file into an array. It then loops over the array and checks the blockchain 20 addresses at a time for the balance. If any address has a balance it sets it aside.

require 'httparty'

tmp_arr = []
addresses_still_with_coins = []
casascius_coin_addresses = []'fulllist.txt').each do |line|
  casascius_coin_addresses << line

casascius_coin_addresses.each_with_index do |address, index|
  tmp_arr << address.gsub("\n","").gsub("\r","")
  if tmp_arr.count == 20
    coin_string = tmp_arr.join(",")
    success = HTTParty.get("{coin_string}")
    success['data'].each do |balance|
      if balance['balance'] > 0
        addresses_still_with_coins << balance['address']
      tmp_arr = []
end"unspent_casascius_coins.txt", "w+") do |f|
  addresses_still_with_coins.each { |element| f.puts(element) }

You might have noticed that this code leaves 19 addresses at the end that haven't been checked. Yea I just plugged those in by hand.

The results can be seen in this gist 'Unspent Casascius Coins'. There were 25101 coins which had a positive balance which would lead me to believe that they had never been spent.

Please let me know if you see any fault in my logic here or if you have a coin(s) on the fulllist that you've not spent which isn't on my final list of 'Unspent Casascius Coins.'


Thanks for reading! Follow me on Twitter and/or G+ for more.

See something wrong or got something to add/change? Fork this git repo and send me a pull request

If you particularly enjoy my work, I appreciate donations given with


27 March 2014