Objective: It is unclear whether the duration of recordings influences estimates of dynamic cerebral autoregulation (dCA). Therefore, we performed a retrospective study of the effects of reducing recording durations on dCA estimates; with the potential to inform recording duration for reliable estimates in challenging clinical populations. Approach: Seventy-eight healthy control subjects and 79 acute ischaemic stroke (AIS) patients were included. Cerebral blood flow velocity was recorded with transcranial Doppler and continuous blood pressure with the Finapres device. The autoregulation index (ARI), derived with transfer function analysis, was calculated for recording durations at one-minute intervals between 1 and 5 minutes using the same starting point of each recording. Main results: Though recording duration did not affect the overall ARI value, when compared to control subjects, AIS patients had significantly lower ARI values for durations between 3 and 5 (p<0.0001), but not 1 and 2 minutes. The intraclass correlation coefficient of all participants, for reproducibility of the five recording durations, was 0.69. AIS patients classified as having impaired cerebral autoregulation (CA; ARI≤4) at 5 min, had a 7.1% rate of false negatives for both 4 and 3 min recordings, reaching 42.9% for 1 min recording. The percentage of false-positives also increased with reduced recording durations (from 0% at 5 to 16.2% at 1 minute). Significance: Reducing recording durations from 5 to 3 min can still provide reliable estimates of ARI, and may facilitate CA studies in potentially medically unstable AIS patients, as well as in other patient groups.