Comparison of Norwegian “linked” decennial census data with statistics compiled from 10 years’ migration registrations showed that the amount of movement omitted by census data was considerable. This hidden movement was of a similar order at every administrative level, but only when total movement was considered. There was wide variation between regions, migration directions, and streams. In some cases census data misrepresented the direction of net movement. Available evidence suggests that these patterns are not confined to Norway, raising the possibility that research findings based on census-derived migration data may merely be artifacts of the data.