OBJECTIVE In mechanically ventilated patients with ARDS, determination of the lower (LIP) and upper (UIP) inflection points of the static pressure-volume curve (P-V) is crucial for planning ventilatory strategies. Recently, a simple new method was proposed for measuring the P-V curve by inflating the lung with constant low flow [14]. We hypothesized that during low flow inflation LIP and UIP might be determined using the pressure-time curve (P-T) instead of P-V. METHODS Eleven paralyzed patients with ARDS were studied. During volume control ventilation the patients were allowed to reach passive functional residual capacity (FRC) and then ventilator frequency, inspiratory to total breath duration ratio and tidal volume (VT) were set to 5 breaths/ min, 80% and 500 or 1,500 ml, respectively. With these settings, constant inspiratory flow (V'I) was administered for 9.6 s and ranged, depending on VT, between 0.05 and 0.15 l/s. P-V and P-T were obtained at two levels of positive end-expiratory pressure (PEEP; 0 and 10 cm H2O), with V'I being achieved either fast (< 0.1 s, minimum delay) or slowly (0.4 s, maximum delay). RESULTS With minimum flow delay for a given experimental condition, the shape of the P-T did not differ from that of P-V. In all cases P-T correctly identified the presence of LIP and UIP, which did not differ significantly between P-T and P-V. With maximum flow delay, compared to P-V, the initial part of P-T was significantly shifted to the left. P-T did not identify the presence of UIP and LIP in one and two cases, respectively. CONCLUSIONS Provided that constant flow is given relatively fast, P-T accurately determines the shape of P-V, as well as the LIP and UIP. Flow delay causes a leftward shift of the initial part of P-T, masking the presence of LIP and UIP in some cases.