Field experiments were conducted to evaluate population densities and survival, developmental rate, and fecundity of grape phylloxera, Daktulosphaira vitifoliae (Fitch), as influenced by root attachment or detachment from mature, field-grown, Vitis vinifera L. grapevines through the growing season. Experiments were performed using artificial infestations of California biotype A grape phylloxera. Thirty-day bioassays on attached- and detached-roots were repeated monthly from May to September in 1997 (cultivar ‘Carignane’) and April to September in 1998 (cultivar ′Thompson Seedless′). The bioassays showed that attached roots had lower population densities than detached roots in all months of both years. Densities varied by month, tending to be higher in spring than in summer. Of the population parameters studied, survival was most influenced by attachment condition, being higher on detached than on attached roots by up to 25-fold. These results imply the importance of vine-related mortality factors to grape phylloxera population density. Influence of root attachment condition on developmental rate and fecundity was not uniform across bioassay months for either year; however, in the four out of 21 assays where there was a significant difference it favored detached roots by twofold. Fruit harvest resulted in higher survival in the July assay but not for assays in August and September; however, neither developmental rate nor fecundity was affected by harvest in any of the assays. We conclude that mortality rather than nutritional factors are most limiting for field populations on susceptible vines. This work suggests that detachment of roots as occurs with root girdling by root pathogens may increase grape phylloxera populations on infested, susceptible vines. These results imply that excised root bioassays over-estimate grape phylloxera virulence and underestimate rootstock resistance.