Using representative members of each of three homologous series of chemicals-ketones, acetates and alcohols-we measured nasal pungency thresholds in anosmics via two stimulus-delivery systems. The first system consists of the fairly commonly used 270 ml, plastic 'squeeze bottles'. The second system consists of 1900 ml, glass vessels with Teflon tubing and nose-pieces. Although bulkier and more susceptible to mechanical breakage, the glass vessels possess advantages that can allow them to provide 'environmentally realistic' chemosensory thresholds, i.e. thresholds closer in absolute values to those that might be obtained under whole-body exposures. Such advantages include a larger volume of the vapor-source to accommodate whole sniffs, and a tight nose-nose-piece connection to avoid stimulus dilution. The outcome revealed that, for every chemical, the glass vessels provided nasal pungency thresholds significantly lower than those provided by the squeeze bottles. The difference amounted, on average, to a factor of 4.6, though the relative potency of the compounds remained the same under both systems. Additionally, when tested with the highest homologues used here, namely, octyl acetate and 1-octanol, anosmics using the glass vessels had little or no difficulty achieving the criterion for threshold whereas they did have difficulty when using the squeeze bottles.