Pastures continue to provide essential functions for the mixed-farming zone in south-eastern Australia, where crop and livestock production are integral parts of most farms. Establishment of pastures in this zone needs to be low-cost and preferably with minimal risk. Pastures are typically sown either directly or in combination with a cover-crop (also called undersowing; the practice of sowing pasture seed simultaneously with a crop that is intended for grain production in the first year), so that the establishment cost is offset by income from the sale of grain. The purposes of this review are to: (i) draw together the literature on undersowing pastures, including studies conducted since the previous review in 1965; (ii) understand why there is a discrepancy between research results that generally do not support the practice of undersowing pastures, whereas farmer preference appears to establish pastures under a cover-crop; and (iii) identify critical needs for further research to aid in making decisions about pasture establishment on-farm. Published and unpublished data from the Australian wheatbelt on establishing pastures by undersowing was examined from the 1920s to the present and included seven publications for perennial species from 26 different experiments. Eight publications addressing establishment of annual species were available from 30 experiments. Many trials appear to have been conducted without being analysed or published. A further 16 international publications were reviewed. Generally, cover-crops reduced annual pasture seedset and perennial density even though the pastures established under the cover-cropping were commonly deemed ‘satisfactory’ by the authors. Pasture establishment was improved by reducing the sowing rate of the cover-crop and/or sowing on alternate rows. Technological change during the past 25 years has led to different configurations of seeding machinery, increased use of herbicides and longer cropping phases. Conclusions previously reached may therefore need to be re-assessed. No data published quantified the risks associated with undersowing on a seasonal and regional basis to determine the probability of success, and there are insufficient data to model the complex physiological interactions between crop and pasture, with most experiments focused on basic agronomic parameters. Furthermore, long-term effects of establishment method on total pasture biomass production and subsequent effects on following crops have rarely been demonstrated, because of the short time-frames in which most experiments have been conducted. Farmers in the mixed-farming zone still prefer to establish pastures under a cover-crop because of perceived financial benefit of the practice. This review identifies substantial research gaps to be addressed to improve pasture-establishment decisions.