The spontaneous breathing trial (SBT) is the final step of weaning from invasive mechanical ventilation. An SBT is aimed at predicting work of breathing (WOB) after extubation and, most importantly, a patient's eligibility for extubation. The optimal SBT modality remains debated. A high-flow oxygen (HFO) has been tested during SBT in clinical study only, which is why no definite conclusion can be drawn on its physiologic effects on the endotracheal tube. Our objective was to assess, on a bench, inspiratory tidal volume (VT), total PEEP, and WOB across 3 different SBT modalities: T-piece, 40 L/min HFO, and 60 L/min HFO. A test lung model was set with 3 conditions of resistance and linear compliance, 3 inspiratory efforts (low, normal, and high), each at 2 breathing frequencies (low and high for 20 and 30 breaths/min, respectively). Pairwise comparisons and a quasi-Poisson generalized linear model that compared SBT modalities were performed. Inspiratory VT, total PEEP, and WOB differed from one SBT modality to another. Inspiratory VT remained higher with the T-piece than in the HFO independent of the mechanical condition, effort intensity, and breathing frequency (P < .001 in each comparison). WOB adjusted by the inspiratory VT was significantly lower during SBT performed with an HFO than when performed with the T-piece (P < .001 in each comparison). The total PEEP was significantly higher in the HFO at 60 L/min than in the other modalities (P < .001). The end points were significantly influenced by breathing frequency, effort intensity, and mechanical condition. With the same effort intensity and breathing frequency, inspiratory VT was higher in the T-piece than in the other modalities. Compared with the T-piece, WOB was significantly lower in the HFO condition and higher flow was a benefit. Based on the results of the present study, the HFO as an SBT modality would seem to require clinical testing.