If they are not related, then I would store them in seperate session
variables.

If they're related....lets say you want to store the
EmpID (int) and EmployeeLoginName (string), I would create a mini wrapper
object.

[Serializable]
public class EmployeeLoginInfo
{
   private int _empID = 0;
   private string _employeeLoginName = string.Empty;

   public EmployeeLoginInfo( int empID , string employeeLoginName )
{
_empID = empID;
_employeeLoginName = employeeLoginName;
}

   public int EmpID { { return _empID;}}
   public string EmployeeLoginName { get { return _employeeLoginName;}}
}

and then I would session that mini object.

Delimited String << Heck No
Array << Heck No

That's my vote.

To me, using the mini object means I only have to pay the casting price
once.
And my code make more sense, because I don't have a bunch of disjoint
properties flying all over the place.

You can check this blog entry as well:

http://sholliday.spaces.live.com/Blog/cns!A68482B9628A842A!151.entry

thanks for the help everyone,
rod.