Fothergill, M., Davies, D. A., Morgan, C. T. (2001). Extensification of grassland use in the Welsh uplands: sheep performance in years 1-6. Grass and Forage Science, 56, (2), 105-117. Sponsorship: UK Ministry of Agriculture Fisheries and Food.An experiment was established in 1991 on a 25-year-old perennial ryegrass/bent (Lolium perenne L./Agrostiscapillaris L.)-dominated pasture in Wales to study the effects of reducing nutrient inputs to previously fertilized upland pasture. The effects of the removal of applications of (1) N (denoted by CaPK) (2) N, P and K (Ca) and (3) N, P, K, and Ca (Nil) were compared with a treatment which received applications of all four nutrients (CaPKN) over a 6-year period (1991?96) in a randomized block design replicated three times. The experiment was managed under a continuous variable stocking regime (ewes and lambs until weaning and ewes thereafter) maintaining a sward surface height of 4?0 cm throughout the grazing season. Although individual liveweight gain of the lambs was unaffected by the treatments, there was a significant reduction (P <0?05) in total lamb liveweight gain, ewe stocking rate and length of grazing season as a result of the withdrawal of nutrients. Over the 6 years total lamb liveweight gain was reduced by 17%, 32% and 45% and ewe stocking rate by 21%, 36% and 49% on treatments CaPK, Ca and Nil, respectively, compared with treatment CaPKN. The effect of withdrawing nutrient inputs on ewe stocking rate was progressive and by 1996 the Nil input treatment displayed a 63% reduction compared with the CaPKN treatment and this was also coupled with a 21-day reduction in length of the grazing season. During the post-weaning period, ewes from the Nil input treatment recorded a liveweight loss in 1995 and only a modest liveweight gain in 1996. This coupled with significantly lower body condition scores (P <0?01) of these ewes in the autumn indicated that the Nil input treatment could lead to reductions in reproductive performance.Peer reviewe