A survey of natural and improved paddocks on 34 farms was conducted over winter within a 60 km radius of Goulburn on the Southern Tablelands of New South Wales. Seventeen of the properties which were found to have significant (greater than 10 per cent) proportions of potentially useful native grasses were resurveyed in the subsequent summer to identify specific Danthonia spp. The survey revealed that useful native yearlong green species such as Microlaena stipoides and Danthonia spp. were relatively abundant in natural paddocks surveyed over winter (16 per cent and 10 per cent respectively), Danthonia being represented by four common species. These natural paddocks, most of which contained Trifolium subterraneum and had some superphosphate input, carried 80 per cent as many stock as their improved counterparts in terms of average DSE/ha. Factors such as cultivation and the presence of naturalized or sown species influenced the abundance of these species, but species differences were apparent. Cultivation reduced the abundance of Danthonia spp. but did not affect the abundance of M. stipoides, which was thus more prevalent than Danthonia spp. in improved pastures. Microlaena stipoides was also more prevalent in uncleared than cleared paddocks. The frequency of Danthonia spp. was negatively associated with high percentage frequencies of annual grasses (principally Vulpia spp.), declining by approximately 2.5 per cent for each 10 per cent increase in annual grasses over winter. The proportion of M. stipoides observed in paddocks appeared to be more influenced by sown pasture species, with lowest percentages recorded where species such as Phalaris aquatica had been sown and highest percentages where no exotic species had been introduced. The abundance of M. stipoides was affected by pH, with percentage frequency increasing as pH declined (down to pH 4 (Cacl~)), indicating the possible acid tolerance of this species. Species' percentages recorded changed with season. For example, Danthonia percentages recorded were considerably greater over summer when flowering heads were obvious than those recorded in winter. This aspect highlights the importance of recording different species during specific seasons to increase the accuracy of frequency estimates.