Fall flaming of harvested peppermint, Mentha piperita L., fields for disease control immediately decreased levels of the predatory phytoseiid mite Neoseiulus fallacis (Garman) by 84–98%, while suppressing the spider mite pest Tetranychus urticae Koch to a lesser extent. By the following spring, more T. urticae were found in flamed fields than unflamed fields and in many cases the pest reached outbreak densities (>5 per leaf) and required use of miticides for control. After harvest on unflamed mint, more N. fallacis were found on prostrate plants near the humid soil surface than on erect plants, whereas spider mites were near equal on prostrate and erect plants. The 1st erect plant type was 15–25 cm tall without leaves on the lower 7.5 cm. The 2nd was 2.5–7.5 cm tall with at least 3 sets of leaves and the prostate plants extended to only 1.3 cm above the soil surface. The net effect of this temporal–spatial asynchrony between predators and prey was to uncouple the system, temporarily. This uncoupling for a short time (physiologically) seemed to stabilize interactions by keeping the predators from overexploiting spider mite populations in the fall and spring. Tests of selective flaming in small field plots showed that predaceous mites dispersed 7–8 m from unflamed plots to recolonize flamed plots before spider mites reached outbreak conditions. Allowing small, unflamed refuges to exist where predaceous mites can survive may reduce negative impact of flaming on control of spider mites at whole field levels.