Site-specific weed management (SSWM) methods allow spatially variable treatment of weed populations according to actual weed abundance, thus offering the opportunity for herbicide savings. However, SSWM's effect on weed population dynamics is not sufficiently understood. In this study, SSWM was conducted based on various application thresholds to analyse the effects on crop yield and weed infestation in the succeeding crop. SSWM was used on a 3.07 ha experimental field in winter wheat (2011) and winter oilseed rape (2012). The whole area was split into application cells of 6 × 10 m and abundance of all weed species was evaluated manually in each cell. Four different herbicide treatments were tested. Standard whole-field herbicide application (blanket spraying) was treatment 1. Treatments 2, 3 and 4 comprised SSWM using different thresholds for post-emergent herbicide application. SSWM resulted in savings of post-emergent herbicides ranging from 71.9% to 100%, depending on the application threshold. Differences in winter rape yield among treatments were generally small and statistically insignificant (P = 0.989). Although some minor changes in weed abundances were observable, the experiment showed that none of the site-specific herbicide treatments caused a significant (α = 0.05) increase of weed species abundance compared to the standard treatment.