Fixed Effects in Stata

Dear All,

I have a panel data with missing observations. I run the following regression:

reg y x

and

tsset id time

xtreg y x

The number of observations is the same in both regressions. My question is: how are missing observations treated in stata? shouldn't the missing values be dropped in the panel analysis if there is less than two observations within each group?

Thanks!