Abstract Hydraulic conductivity of the xylem is computed as the quotient of mass flow rate and pressure gradient. Measurements on excised plant stems can be difficult to interpret because of time-dependent reductions in flow rate, and because of variable degrees of embolism. Using Acer saccharum Marsh. stems, we found that certain perfusing solutions including dilute fixatives (e.g. 0.05% formaldehyde) and acids with pH below 3 (e.g. 10 mol m−3 oxalic) prevent long-term decline in conductivity. Xylem embolism can be quantified by expressing the initial conductivity as a percentage of the maximum obtained after flow-impeding air emboli have been removed by repeated high-pressure (175 kPa) flushes. Correlation between microbial contamination and declining conductivity suggests that long-term (> 4h) declines are caused by microbial growth within the vessels. Unpredictable trends in short-term (< 4h) measurements may be caused by movements of air emboli in vessels and/or participate matter.