Present study was designed to verify which or if any of plastome loci is a hotspot region for mutations and hence might be useful for molecular species identification in feather grasses. 21 newly sequenced complete plastid genomes representing 19 taxa from the genus of Stipa were analyzed in search of the most variable and the most discriminative loci within Stipa. The results showed that the problem with selecting a good barcode locus for feather grasses lies in the very low level of genetic diversity within its plastome. None of the single chloroplast loci is polymorphic enough to play a role of a barcode or a phylogenetic marker for Stipa. The biggest number of taxa was successfully identified by the analysis of 600 bp long DNA fragment comprising a part of rbcL gene, the complete rbcL-rpl23 spacer and a part of rpl23 gene. The effectiveness of multi-locus barcode composed of six best-performing loci for Stipa (ndhH, rpl23, ndhF-rpl32, rpl32-ccsA, psbK-psbI and petA-psbJ) didn't reach 70% of analyzed taxa. The analysis of complete plastome sequences as a super-barcode for Stipa although much more effective, still didn't allow for discrimination of all the analyzed taxa of feather grasses.