Few studies have evaluated seed quality in native to semi-arid zones' grasses. The objective was to validate published methodologies to carry out viability tests using Tetrazolium (TZ) test for caryopses (Cs) of Bouteloua curtipendula , Leptochloa dubia , Digitaria californica , Setaria macrostachya and Pennisetum ciliare , considering the imbibition time, TZ concentration and cut in the Cs. In addition, showing staining patterns according to evaluated treatments. Viability and vigor data was expressed as percent and hence transformed to arcsine. A completely randomized experimental design was carried out using a factorial arrangement: A) imbibition time before cutting; B), type of cut; C), TZ concentration, and Tukey's test (p<0.05) for mean comparisons. The Cs longitudinal cut was the most adequate (P<0.05), it allows observing Cs' embryonic structures and determining damaged sites; however, the Cs of D. californica and S. machrostachya are permeable to TZ and it was not necessary to section the Cs, which facilitates its evaluation. The ISTA protocols for viability in Poaceae are not directly applicable for the species evaluated, because the imbibition time for these species is not considered, which was 3 to 4 hours. It is recommended to use a 0.5% TZ concentration and evaluate 12 hours after the imbibition time between Cs and the TZ solution.