During 1980 and 1981, a large-scale trial of the biological control of T. urticae Koch on protected strawberries, using the predatory mite P. persimilis Athias-Henriot, was done in 52 "walk-in" plastic tunnels on a commercial nursery in southern England [UK]. Introductions of predatory mites in March or early April at a rate of one per plant were consistently successful. In 1980, when predators had not been introduced the previous year (so enabling large numbers of spider mites to overwinter) peak mite populations were large, causing slight foliar damage in some tunnels before control was achieved. In 1981, only a small residual population of spider mites had survived from the previous year and peak numbers were small; the crop appeared healthy and there was no visible damage by mites. For adequate control, reintroduction of predators was found to be necessary when crops were replanted in the summer following soil cultivation and sterilization with methyl bromide. A population assessment method for two-spotted spider mite based on counting aggregations of the mites was devised. The mean number of spider mites (all stages) per leaf was found to be approximately the same as the number of leaves with 5 or more mites present in a 25-leaf sample.