Monitoring programs that assess species-richness and turnover are now regarded as essential to document biodiversity loss worldwide. Implementation of such programs is impeded by a general decrease in the number of skilled naturalists. Here we studied how morphotypes, instead of species, might be used by unskilled participants (referred to as “volunteers”) to survey common plant communities. Our main questions were: (1) Can morphotypes be used as a robust estimator of species-richness (alpha-diversity) and assemblage turnover (Beta-diversity)? and (2) What is the robustness (reproducibility and repeatability) of such methods? Double inventories were performed on 150 plots in arable Weld margins, one by a non-expert using morphotypes, the other by a taxonomist using species. To test the robustness of morphotype identiWcation among participants, 20 additional plots were surveyed by eight volunteers using the same protocol. We showed that (1) the number of morphotypes identiWed by unskilled volunteers in a plot was always strongly correlated with species-richness. (2) Morphotypes were sensitive to diVerences among habitats but were less accurate than species to detect these diVerences. (3) Morphotype identiWcation varied signiWcantly within and between volunteers. Due to this lack of repeatability and reproducibility, parataxonomy cannot be considered a good surrogate for taxonomy. Nevertheless, assuming that morphotypes are identiWed with standardized methods, and that results are used only to evaluate gross species-richness but not species turnover, parataxonomy might be a valuable tool for rapid biodiversity assessment of common wild flora.