Pteridium aquilinum is a problematic, perennial, invasive species worldwide that poses serious problems in the British uplands. However, there is a lack of knowledge on long-term success in terms of weed control and land improvement. We assessed the effects of six P. aquilinum-control treatments at two acid-grassland sites in the Scottish Borders (Sourhope 1 & 2). There were six treatments: (i) untreated, (ii) cutting once-yearly, (iii) cutting twice-yearly, (iv) asulam sprayed in year 1, (v) cutting once in year 1/asulam in year 2, and (vi) asulam in year 1/cut in year 2. The annual cutting treatments were stopped after 9/10 years. We measured the response of three variables over 25/26-years assessing P. aquilinum-performance, agricultural-improvement and species richness. We estimated the treatment “effect window”, defined as the period over which the P. aquilinum-control treatments had a significant effect relative to untreated experimental-controls. Results were completely different for all variables in the two experiments. At Sourhope 1 the cutting treatments had much longer “effect windows” than the herbicide-based treatments in reducing P. aquilinum performance, especially the cut twice-yearly treatment with a 25-year window. At Sourhope 2, the treatment differences were much reduced with the best treatment (cut twice yearly) producing a 13-year window. At both experiments, asulam-based treatments produce a minimum “effect window” of 10 years. Increases in both the agricultural-improvement index and species richness were also obtained at both sites. “Effect windows” for both agricultural-improvement index (TVI) and species richness also showed some improvement; “effect windows” were obtained of 12–20 and 9–20 years for TVI, and 7–17 and 2–10 years for species richness at Sourhope 1 and Sourhope 2 respectively. Species richness declined though time in both experiments. Cutting, and preferably twice per year, was the most effective treatment but also the costliest, with the most cost-effective approach being a single asulam spray. However, the length of the “effect windows” suggest that repeat-spraying every 10-years is needed. Unfortunately, where the aim is to recover biodiversity interventionist techniques such as seeding will also be needed. The use of “effect windows” for estimating long-term impact of perennial weed control is novel and may be of use in other situations.