Please explain what you are trying to test for.  Since a variable can have a different value on each observation in a dataset what do you mean by a blank variable?  Do you want to know if it ever is blank? Or if it always is blank? What do you mean by blank?  Are you only testing character variables?  What about numeric variables?  Do you want to test for missing values?  What about special missing values?   If your variables do not have too many levels then a quick way to test if there are any missing (and any non-missing) values is to use the NLEVELS option on PROC FREQ.     proc freq data=have nlevels ; tables varname / noprint; run; ... View more

Ok, here you go.  0 indicates no missing data items, >0 is count of missings: data abcd; input memname $ name $ libname $; status2=0; cards; class age work class height work ; run; data class; set sashelp.class; if _n_ in (3,4) then age =.; run; data _null_; set abcd; call execute(cat('proc sql; update abcd set status2=(select sum(case when ',strip(name),' is null then 1 else 0 end) from ',catx('.',libname,memname),') where libname="',strip(libname),'" and memname="',strip(memname),'" and name="',strip(name),'"; quit;')); run; ... View more

Or   Proc print data=<any valid data set name>; run;   Will print the entire data set, all values all rows.   Or you need to show by concrete example what you mean for output.   Once you have created a data set the only way to tell "new" variables, if this is the core of your question and not printing, then you would have to preserve a list of variables from the "old" data set. ... View more