Three field experiments showed that perennial ryegrass can be successfully slot-seeded into an existing Festuca-Agrostis sward, Ryegrass survival was satisfactory in all cases but was improved on a low fertility site by the application of fertilizer N, P and K in the slot at sowing.The effect of varying inter-row spacing from 37-5 to 15 cm was measured in a 3-year experiment. In year one, involving seven cuts, slot-seeding increased total herbage dry matter harvested by a mean value of 17%; inter-row distances of 22-5-30 cm gave the optimal combination of ryegrass -I-old sward herbage. Differences in yield between inter-row spacings declined in the subsequent two years, as the rows of ryegrass thickened, Ryegrass digestibility (measured only in year two, from six cuts) was higher than that of the old sward; total metabolizable energy harvested from 15-cm rows was 20% higher than that from unsown controls. Increasing N input from 200 to 400 kg ha~', starting in year two, only produced a significant increase in total herbage harvested in year three but proved effective in increasing the ryegrass contribution from the wider spaced rows.