Studies investigating urbanisation effects on pollinator diversity have shown contrasting results, ranging from predicting urban areas as ecological sinks to possible refuges for pollinators. These divergent results may stem from the diversity of non-urban areas being compared with urban areas as well as the existence of interactions among habitats. We tested the interactive effects of urban, agricultural and semi-natural areas on the richness of flower-visiting insects at the landscape scale. Using the dataset from a citizen science programme, we investigated how the richness pattern varied (i) along the triple gradient across urban, agricultural and semi-natural areas and (ii) among the different visitor taxa. We stratified the dataset to have a balanced landscape composition and spatial distribution of sampling sites for each plant species sampled. Accounting for general sampling factors (e.g., plant species, temperature, shading, time of observation), we showed that urbanisation negatively impacted the total richness of flower visitors, especially when associated with agricultural areas. We observed maximum richness in agricultural landscapes composed of ∼30 % of semi-natural areas, thus suggesting synergistic effects between these land uses on insect richness. Our results showed contrasting richness patterns for different taxa. For bees, urban areas were less detrimental to richness than agricultural areas. By contrast, for syrphids and butterflies, richness was highest in agricultural and semi-natural areas, respectively. Our study revealed the necessity to distinguish among rural areas when studying urbanisation gradients and highlighted synergistic landscape effects that could help inform landscape planners about how to attract a diversity of flower visitors.