A new approach to calibration of the Australian water balance model (AWBM) daily rainfall-runoff model was tested by transposing calibrated parameter values among 18 catchments in coastal Queensland, Australia. The calibration procedure adjusts input rainfall and evaporation by linear scaling with constraints on the calibrated value of average surface storage capacity to maximize the coefficients of efficiency based on monthly totals of runoff. The calibrated parameter values on each catchment were able to estimate runoff on each of the other 17 catchments with an average error of only 5.5%. Some 94% of the 306 (18ױ7) estimates of runoff were within 15% of actual runoff. Some 71% of the coefficients of efficiency from the tests of transposing parameter values were >=0.8. A second stage of the study was to average the 18 sets of calibrated parameter values to a single set that estimated runoff from the 18 catchments with an average error of only 5.3% and a maximum error of 15.8%. The new method of calibrating the AWBM offers potential for estimating runoff from ungauged catchments with small errors.