In this paper, we evaluated the influence of some glass ampoules characteristics that are critical as far as secondary particle contamination (visible and sub-visible) of small volume parenterals. Following characteristics were considered: snap-opening system (colour-break scoring and o.p.c.); break-point strength (high and low); dimensional characteristics (three types D1, D2, D3). According to the proper statistical design, 18 different batches of 2 ml ampoules filled with water for injection were manufactured; then sub-visible and visible particulate contamination generated on snap-opening the ampoules were examined. Sub visible particle counting was performed by the light-blockage method and visible particles counting by the membrane-filtration method. With regard to sub-visible particles, statistical analysis showed significant interactions of break-point strength with both ampoule dimensional characteristics and the type of snap-opening system. Break-point strength seems to have the greatest influence on the level of particulate contamination. Differences in ampoule snap-opening systems and dimensional characteristics exert a significant effect on the level of particle contamination only at high break-point strength. With regard to visible particles generated on snapping the ampoules, data are consistent with the findings of sub-visible particulate contamination. Ampoules manufactured according to DIN dimensional characteristics and with an o.p.c. snap-opening system provide the lowest particle burden.