The cerebral microvasculature is exquisitely sensitive to carbon dioxide (CO2), and cerebrovascular reactivity to CO2 is commonly quantified as a measure of cerebrovascular function. However, the details of cerebrovascular reactivity protocols and analyses vary considerably between research groups. We aimed to examine whether these different methodologies and analyses influence the results of cerebrovascular reactivity testing.We used dynamic end-tidal forcing to measure cerebral reactivity to CO2 in healthy humans (n=7; 4 male) under conditions of normoxia (PETO2=100mmHg), hypoxia (PETO2=50mmHg), and hyperoxia (PETO2=150mmHg). Progressive hypocapnia was obtained by voluntary hyperventilation to a target PETCO2=20mmHg, and hypercapnia by hypoventilation with increased inspired CO2 until PETCO2=55mmHg was achieved. Supine beat-to-beat blood pressure (Finometer) and middle cerebral artery blood flow velocity (MCAv; transcranial Doppler ultrasound) were continuously recorded, as were breath-by-breath PETCO2 and PETO2. Values are mean±SEM, compared by two-way repeated measures ANOVA.The sensitivity of the response is reported as the gradient of the relationship between MCAv and PETCO2 based on a linear, segmented linear, or sigmoidal fit. There was a significant effect of the oxygen condition on the sensitivity (P=0.012). There was a significant effect of the curve fit on the sensitivity (P<0.001). There was no significant interaction between condition and curve fit (P=0.156) (Figure 1). Comparisons of the Akaike information criterion (AIC) derived for each fit revealed a significantly worse fit across all conditions for the linear regression (linear: 327±14; segmented linear: 313±15; and sigmoid 315±14, P<0.05). The estimated information loss was equivalent for segmented linear and sigmoid fits. In some studies, control of PETO2 is initiated only after completion of hyperventilation to the target nadir for PETCO2. We evaluated whether this impacted our results. There were no significant differences in the slopes or AIC across conditions between early or late PETO2 forcing. However, when forcing was initiated after hyperventilation there were no significant differences in reactivity between conditions regardless of the fit applied. The sensitivity was consistently larger across conditions when a sigmoid fit was applied. Based on these analyses, the choice of curve fitting may influence the outcome of cerebral reactivity testing. Considering the ability to detect a difference between conditions, rationale for curve fit based on the known response characteristics, and AIC comparison, a sigmoidal fit is recommended. Initiating forcing of PETO2 after hyperventilation may influence the ability to detect changes in cerebral reactivity during different PETO2 conditions.