Suppose you have a simple circuit with a voltage source V1 connected to a resistor R1, like this:
You could connect an ammeter in series, and then the internal resistance of the ammeter would affect the actual current reading, introducing some error. But you could also connect a voltmeter (with a high internal resistance) in parallel across R1, and calculate current by dividing the measured voltage by R1. There would still be some error due to the voltmeter's internal resistance, but which would be more accurate? Or more specifically, under what conditions (i.e. large/small current, R1, V1, etc.) would it more accurate to use the second approach, with a voltmeter instead of an ammeter?