Weekly atmospheric deposition samples were collected from early May to early September during 1984 to 1987 in a small agricultural watershed near Regina, Saskatchewan, an area in the midst of extensive agricultural pesticide use in the Canadian prairies. “Bulk” deposition samples were collected weekly in duplicate and analyzed for six herbicides (bromoxynil, 2,4–D, dicamba, diclofop, triallate, and trifluralin) and five insecticides (carbaryl, carbofuran, chlorpyrifos, deltamethrin, and dimethoate). In general, no significant differences were found among the duplicate samples. Over the 4‐year period, the frequency of weekly detection of herbicides during the growing seasons was 0, 32, 39, 53, 58, and 67% for trifluralin, bromoxynil, triallate, diclofop, dicamba, and 2,4–D, respectively. For insecticides, only carbaryl and carbofuran were detected in one to two samples in the years for which the insecticides were analyzed. This infrequent detection reflected the relatively high detection limits set for the insecticides as compared to those for the herbicides. The seasonal herbicide atmospheric deposits during the growing period over the 4‐year period (mg ha−1) ranged from 90 to 234 (bromoxynil), 386 to 1,374 (2,4‐D), 20 to 219 (dicamba), 43 to 338 (diclofop), and 9 to 260 (triallate). Highest deposition rates were measured during and immediately following the spraying periods, with lesser, but significant, amounts also being detected during the harvest operations in August.