ABSTRACT This study investigated the suitability of honey sample collection methods for determining the botanical origin of honey through palynological analysis. We used three methods to collect honey samples in three different modes viz. extracted honey using a honey extractor, squeezed honey and pipetted honey (collected by micropipette/dropper from honey cells only) during 2017 to 2019 in West Bengal, India. We considered two native honey bee species (Apis dorsata and Apis florea) and one introduced bee species (Apis mellifera). Pollen composition differed significantly, both quantitatively and qualitatively, among the honey samples of the different methods. The number of pollen grains in extracted honey and squeezed honey was significantly higher than that of pipetted honey. Furthermore, some pollen types of nectar deficient, but polleniferous plants (viz. Capparis zeylanica, Echinochloa frumentacea, Papaver somniferum, Poaceae type, Nelumbo nucifera, Solanum melongena, and Solanum sisymbriifolium), were also present in extracted and squeezed honeys. We concluded that some pollen grains present in extracted and squeezed honey samples came from stored pollen loads or bee bread in the hive. Hence, the pollen spectrum for pipetted honey samples was more accurate in depicting the bees foraging on nectariferous plants.