Coin finds are used to inform discussion about coins in circulation and, hence, economic activity. This discussion often rests on the assumption that patterns in finds of accidentally lost coins mirror to a useful extent those of the coins in circulation at the time. Whenever possible, such assumptions should be tested. This study does this empirically with data from a known coinage. Conditional support was found for the assumption. There is also some evidence that coin size and denomination can affect the coin record but these variables did not add significantly to predictions about coins in circulation based only on the numbers of coins found. Some conditions and precautions are suggested when using the assumption and some uses of data based on accidentally lost coins are discussed.
