In Davis, Calif., cabbage aphids, Brevicoryne brassicae (L.), and green peach aphids, Myzus persicae (Sulzer), were sampled from collards planted October 1982 and March 1983 in replicated plots (3 by 7 m) where weeds were subjected to one of three treatments--mowed, tilled, or unmanaged. In the October planting, collards and weeds began growth simultaneously, and weeds had no apparent impact on collard growth or on colonization by either aphid species. Parasitism of aphids by Diaeretiella rapae (McIntosh) was significantly higher in plots containing unmanaged weeds only once (6 December), and during early November secondary parasitism, mostly by Alloxysta fuscicornis (Hartig), was greater in tilled and mowed plots (18-38%) than where weeds were unmanaged (0-5%). In the March planting, collards were planted among preexisting weeds; collard growth among weeds was retarded as was the development of aphid populations. Primary parasitism was negligible (as were aphids) in unmanaged weedly plots and greater in mowed (7.2%) than in tilled plots (3.6%). The overall impact of parasitism on aphid populations was minimal. Secondary parasitism (again mostly by A. fuscicornis) averaged 40.8% in tilled, 7.5% in mowed, and none in weedy plots.