Heavy-metal fluoride glasses are very promising optical fiber materials because of their predicted ultralow loss and long transparency range. Although conventional silica fibers have attained their theoretical minimum loss of 0.15 dB/km, fluoride glasses have the potential to yield losses of only 0.001 dB/km. Fluoride glasses also exhibit transparency into mid-IR frequencies, a region inaccessible to silica fibers. However, this group of glasses is very unstable to devitrification during both bulk glass synthesis and fiber-drawing. This instability has limited their commercial exploitation to a small niche market in the laser industry. The ZBLAN glass (53ZrF(4)-20BaF(2)-4LaF(3)-3AlF(3)-20NaF) is the most promising of these materials since its fiber-drawing region lies on the edge, or possibly just outside its crystallization region. It is believed that additional research into understanding the nucleation mechanics involved in the devitrification of fluoride glasses will lead to the development of technology to suppress such nucleation, or at least minimize the associated crystallization temperature region, allowing high optical quality fibers to be produced. It has recently been demonstrated that a microgravity environment can suppress devitrification in ZBLAN glass preform preparation, and that devitrification may be reduced when preparing ZBLAN terrestrially in a containerless facility. It is believed that the role of viscosity is critical in the devitrification mechanism of ZBLAN glass and in determining the optimum fiber-drawing temperature. Unfortunately, viscosity data for fluoride glasses are only available above the melting point and around the glass transition. A piezoelectric viscometer has been developed and is being used to determine the missing viscosity data in the fiber-drawing and crystallization temperature regions. Shear thinning of the glasses and/or the application of hydrostatic pressure on the glasses have been recently proposed to be responsible for devitrification during fiber-drawing at 1 g and in reduced gravity. The study we report here is to explore the extent to which such a proposal is realistic.