ABSTRACT: A case study based on a wholly agglutinated Nothia-dominated assemblage from Santonian flysch strata of the Silesian Unit, Western Carpathians, shows that standard counting including each fragment of tubular taxa as a specimen leads to extreme exaggeration of their relative abundance. A method of defragmentation of fragmented foraminiferal specimens is here proposed to address this problem. Following the method we can recalculate test fragments into individuals based on comparison with complete tests known from the fossil record or modern analogs. Application of this method on the Nothia-dominated flysch type assemblage from the Carpathians turned it into a Karrerulina - Recurvoides assemblage with Nothia as a scarce element. The proportion of Nothia decreased from 46% to les then 1% and all tubular agglutinated taxa from 74% to 4%. At the same time, the defragmentation substantially increased the Shannon diversity index, equitability, and decreased the dominance in the assemblage. Defragmentation of flysch-type assemblages also decreases the faunal density (abundance) of flysch-type assemblages. A comparison of fragmentary material with complete specimens revealed that a single individual of Nothia sp. 1 produces a few hundred fragments and Rhizammina many hundreds. Simple tubular forms like Bathysiphon, Hyperammina and Saccorhiza break into fewer fragments (5 to 23). The small share of tubular agglutinated taxa in flysch-type assemblage demonstrated by the case study does not mean their small importance. They are an important trophic indicator and they may still be a dominant component in the sense of biomass. They should be treated, however, as a qualitative feature based on their presence rather then as a quantitative feature based on their share of the assemblage. A comparison of the studied assemblage after defragmentation with other wholly agglutinated assemblages from the Carpathians shows just a minor difference between lower slope and abyssal faunas in terms of taxonomic composition and diversity. Exaggeration of tubular taxa share in quantitative analysis should be avoided for more reliable comparisons of sub-CCD assemblages. In applied research, exclusion of tubular fragments from quantitative analysis may be a practical solution instead of time-demanding defragmentation. A count of defragmented assemblage does not differ substantially from a count excluding the tubular taxa. Amore accurate method is establishing the number of fragments per complete test by defragmentation of a typical sample and application of results to estimate the number of individuals.