Paterson's curse and related weeds (Echiurn spp.) were introduced as garden flowers before 1850 and have spread to over 30 million ha in southern Australia. Four hundred successful releases of crown weevil (M. larvatus) populations specifically targeting Echiurn spp. were made in the 1993-2000 period. Based on the timing, location and performance of these past releases of beneficial insects, spatially and temporally specific trajectories of biocontrol have been simulated. Insect populations established by the past releases are expected to cover expanding areas at densities sufficient to limit host Echiurn infestations only over the next 25-50 years. The present analysis tackles the questions of where and how many additional releases are economically justified to speed up this process. We identify 31 districts in which diminishing niches for further insect releases are projected over time, according to the locations of damaging weed infestations and the timing, location and numbers of past insect releases. Benefits of biocontrol are expressed in terms of the value of recovered pasture productivity, keyed to estimates of loss and to historical district livestock inventories converted to dry sheep equivalent (DSE) feed availability levels to which prices are applied. Expected marginal contributions of increments of new releases were simulated for each of the 31 districts, subject to the space/time limitations of each niche. Our explicit accounting for the spatial and temporal dimensions has made possible the economically optimal targeting of new biocontrol releases. For example, at $12/DSE and a marginal cost of $2000 per release, with a discount rate of 10%, we find there is a case for a program of over 400 new releases targeted to 17 districts, with as few as five releases to each of several and as many as 70 releases in one district. Crown Copyright© 2002 Published by Elsevier Science B.V. All rights reserved.