Failure over the past two to three decades to implement industry-led, systematic forage evaluation systems that translate forage performance data to animal production and economics means that the livestock industries are poorly positioned to judge how much economic benefit they are gaining from forage plant improvement and to propose future priorities and targets. The present paper identifies several knowledge gaps that must be filled to enable the value being delivered to pasture-based livestock industries by forage improvement to be determined, demonstrated to farmers and increased in the future. Seasonal yield, total annual yield, nutritive value and feeding value of pasture are all important traits for driving the productivity of pasture-based livestock production systems. From a farm systems perspective, persistence of the yield or quality advantage of new cultivars is also economically important. However, this is the least well defined of the productivity traits considered in the paper. Contrary to anecdotal reports, evidence indicates that the genetic potential of modern ryegrass cultivars to survive in grazed pastures is at least equivalent to that of older cultivars. Plant breeding in Europe and New Zealand has changed the seasonal yield, quality and intake potential of perennial ryegrass. On the basis of dry matter (DM) yield data from small-plot evaluation trials, the New Zealand forage value index indicates that the top-ranked perennial ryegrass cultivars offer between NZ$280 and NZ $650/ha per year potential additional operating profit to dairy businesses (depending on region), compared with a historical genetic base of cultivars that were first entered into yield testing programs between 1991 and 1996. The equivalent figure in Ireland (including nutritive value effects) is about €325/ha per year. These estimates are yet to be confirmed in animal production studies. In intensive dairy systems, current rates of genetic gain in DM yield lag well behind realised rates of gain in animal genetics and associated increases in feed demand per animal. Genetic gains in yield need to double from current rates (estimated at 0.5% per year); but, it is not known whether this is possible in an outcrossing species such as perennial ryegrass, which is normally grown in a mixture with other species, especially white clover. Improvements in DM yield in seasons where extra DM has greatest economic value in grazing systems should dominate breeding objectives, but this must now be augmented by consideration of the environmental impacts of intensive pasture-based livestock production systems and opportunities to mitigate this through germplasm selection. There is less evidence that nutritive or feeding value of ryegrass cultivars significantly limits animal production and profitability and useful improvements have already been made using tetraploids and later heading material.