Background: In patients with acute ischaemic stroke, hyperglycaemia has been retrospectively associated with negative outcome. There is an ongoing discussion as to which treatment algorithm, if any, provides the most effective prospective intervention. Here we test the safety and feasibility of an intravenous insulin-only infusion protocol designed for pragmatic routine clinical use. Methods: 40 ischaemic stroke patients with onset <24 h ago, admitted to our stroke unit, were randomized either to the study regimen (50 IU insulin in 50 ml 0.9% saline solution applied intravenously via a perfusor pump), with the aim of reaching and maintaining blood glucose levels between 4.44 mmol/l (80 mg/dl) and 6.11 mmol/l (110 mg/dl), or were treated with insulin subcutaneously if concentrations were above 11.10 mmol/l (200 mg/dl). Treatment was continued for 5 days. Primary outcome was the number of hypoglycaemic (<3.33 mmol/l; <60 mg/dl) and severe hyperglycaemic (>16.65 mmol/l; >300 mg/dl) events. Results: Hypoglycaemic events were significantly more common in patients treated intensively (total n = 25; incidence rate ratio, IRR = 5.3; 95% CI = 1.2–22.4; p < 0.05). Symptomatic events were rare (total n = 5). Severe hyperglycaemia was associated with conventional treatment (IRR = 4.9; 95% CI = 1.5–15.9; p < 0.05). Though those treated intensively attained near-normoglycaemic levels quicker and had significantly lower blood glucose levels over the study period (6.49 ± 2.19 mmol/l vs. 8.01 ± 3.06 mmol/l; 95% CI = –1.78 to –1.28, p < 0.0005), treatment imposes considerable strain on both patients and caregivers. Conclusions: The intensive intravenous insulin infusion protocol effectively lowers blood glucose levels with an increased risk of manageable hypoglycaemic events. However, a highly motivated and trained staff seems essential, limiting feasibility outside of specialty care settings.