A filter–paper technique that can be used to measure the matric potential of field soil is described in detail. When a calibrated batch of filter papers is used, the precision is limited by the variability between individual papers and the limitation of weighing to the nearest mg. The absolute accuracy is also limited by the instruments used for calibration. There was reasonable agreement between batches of the same type of paper, suggesting that our generalized calibration curves can be used for Whatman No. 42 filter papers where accuracy is not paramount.

For potentials ≥−2.5 MPa, papers need to be equilibrated with the soil for 6 d. To achieve the same accuracy at lower potentials longer equilibration periods are required. Results suggest that the technique should be effective down to at least −100 MPa if soils are allowed to equilibrate fully with the papers. Coefficients of determination (r2) for the calibration lines were all ≥0.92. A regression of log(-matric potential) measured by the filter paper technique against measurements made using tensiometers and a psychrometer over a range of potentials between −1 kPa and −10 MPa gave an r2 value of 0.995.