Xylem hydraulic conductivity (K) is typically defined as K = F/(P/L), where F is the flow rate through a xylem segment associated with an applied pressure gradient (P/L) along the segment. This definition assumes a linear flowpressure relationship with a flow intercept (F0) of zero. While linearity is typically the case, there is often a non-zero F0 that persists in the absence of leaks or evaporation and is caused by passive uptake of water by the sample. In this study, we determined the consequences of failing to account for non-zero F0 for both K measurements and the use of K to estimate the vulnerability to xylem cavitation. We generated vulnerability curves for olive root samples (Olea europaea) by the centrifuge technique, measuring a maximally accurate reference Kref as the slope of a four-point F vs P/L relationship. The Kref was compared with three more rapid ways of estimating K. When F0 was assumed to be zero, K was significantly under-estimated (average of -81.4 +/- 4.7%), especially when Kref was low. Vulnerability curves derived from these under-estimated K values overestimated the vulnerability to cavitation. When non-zero F0 was taken into account, whether it was measured or estimated, more accurate K values (relative to Kref) were obtained, and vulnerability curves indicated greater resistance to cavitation. We recommend accounting for non-zero F0 for obtaining accurate estimates of K and cavitation resistance in hydraulic studies.