Recently, the use of Pre-Crossconnected Trails (PXTs) has been proposed for fault tolerant WDM optical networks. Arranging protection bandwidth into PXTs eliminates the need for switch reconfiguration after a failure which leads to faster recovery times. The researchers who introduced PXTs gave experimental results showing that PXT protection performs not much worse and sometimes better than conventional algorithms in terms of bandwidth efficiency. Their experiments compared a greedy PXT algorithm with a simple non-greedy conventional algorithm, so there remained some question about the effect of the PXT constraint itself on bandwidth efficiency. In this paper, we explore this topic further. We use experimental results to compare simple and greedy heuristics both with and without the PXT constraint. Our results show that the PXT variations require significantly more bandwidth overhead than their conventional counterparts especially in larger networks.