Pollination is an ecological service with high economic values related to human and ecological wellbeing. This study aims to evaluate the correlation between landscape complexity and pollination service (bee) in Qazvin province at the landscape scale. Landscape pattern metrics (Number of Patches (NP), Patch Richness (PR), Edge Density (ED), and Shannon Index (H)), as a proxy of landscape complexity, were calculated by the model builder in the ArcGIS and pollination service was generated by InVEST. The correlation between landscape complexity and pollination service was evaluated by SPSS software. Results of the correlation between landscape metrics and pollination service, using Linear regression, showed that H has the highest and the most significant relationship with pollination at the 0.01 level (0.420**). Correlation between other metrics and pollination was found to better explauned by Quadratic regression, showing a significant correlation with NP (0.171**) and PR (0.352**) at the 0.01 level and a non-significant relationship with ED (0.002). It was concluded from our results that in the low NP and PR, correlation is high and positive, and in the high NP and PR, correlation is negative, indicating the effect of fragmentation on pollination service. The results of this study can be used in agricultural system planning and biodiversity management to preserve the ecological processes.