The absence of a BATSE line detection in a gamma-ray burst spectrum during the mission's first six years has led to a statistical analysis of the occurrence of lines in the BATSE burst database; this statistical analysis will still be relevant if lines are detected. We review our methodology, and present new simulations of line detectability as a function of the line parameters. We also discuss the calculation of the number of ``trials'' in the BATSE database, which is necessary for our line detection criteria.