Pat, I checked Sullivan County, 1850 census just now (I'm also researching Rogers in Sullivan County -- although different family). I only see two daughters named Sarah ... one is only 3 so unless your age is off I doubt this is her. This is Sarah Rogers, aged 3, daughter of Greenberry Rogers. The other is more interesting and not solid fact. There is a family headed by Nicholas Rogers... Nicholas's family spans a page and when you jump to the second page, the census taker usually rewrites the surname of the family. In this case, he wrote Booher (not Rogers) which happens to be the name of the family just in front of Nicholas Rogers. After Sarah "Booher" are two more of their children (Reuben and Ellen). The implication here would be they are Boohers too since he wrote that by Sarah... but in 1860, they are Rogers. So... while the proof isn't conclusive (and this Sarah was born 1840)... I'd say this is a good clue to focus on. It would be very unusual for a stepchild to be listed in the middle of a family group without relisting the family name for the next child which at least gives you a reason to think Booher was a census taker mistake (ie., he should have written Rogers but simply looked at the previous page wrong) ... and it gives you Nicholas Rogers as a possible father to focus on... Gary