Artificial light at night (ALAN) is a relatively new and rapidly increasing global change driver. While evidence on adverse effects of ALAN for biodiversity and ecosystem functioning is increasing, little is known on the spatial extent of its effects. We therefore tested whether ALAN can affect ecosystem functioning in areas adjacent to directly illuminated areas. We exposed two phytometer species to three different treatments of ALAN (sites directly illuminated, sites adjacent to directly illuminated sites, control sites without illumination), and we measured its effect on the reproductive output of both plant species. Furthermore, in one of the two plant species, we quantified pre-dispersal seed predation and the resulting relative reproductive output. Finally, under controlled condition in the laboratory, we assessed flower visitation and oviposition of the main seed predator in relation to light intensity. There was a trend for reduced reproductive output of one of the two plant species on directly illuminated sites, but not of the other. Compared to dark control sites, seed predation was significantly increased on dark sites adjacent to illuminated sites, which resulted in a significantly reduced relative reproductive output. Finally, in the laboratory, the main seed predator flew away from the light source to interact with its host plant in the darkest area available, which might explain the results found in the field. We conclude that ALAN can also affect ecosystem functioning in areas not directly illuminated, thereby having ecological consequences at a much larger scale than previously thought.