Very little is known about the rate at which pollen grains are mobilized within insect-pollinated crop systems, and this is especially true the for commercial production of field-grown cucumber (Cucumis sativus L.), monoecious muskmelon (Cucumis melo L.), and triploid watermelon [Citrullus lanatus (Thunb.) Matsum. & Nakai]. The rates of pollen depletion for these crops were therefore investigated on plots simulating commercial crop production using a mixed honey bee (Apis mellifera L.) and bumble bee (Bombus impatiens Cresson) pollinator complex. At anthesis, staminate cucumber, muskmelon, and watermelon flowers contained on average 10539, 11176, and 30739 pollen grains/flower, respectively. At the time flowers closed in the early afternoon (1300 to 1400 hr), only 61% of the total pollen produced had been removed from staminate cucumber flowers, 44% to 62% from muskmelon, and 81% from watermelon flowers. The results suggest that total pollen production in these crops may not necessarily reflect total pollen availability to floral visitors (bees). However, of the total amount of pollen actually removed per flower, >57% occurred during the 2 h following flower anthesis of cucumber and muskmelon, and >77% occurred during the 2 h following flower anthesis of watermelon. Thus, most of the accessible pollen was removed shortly after anthesis, which is when these crops are most receptive to pollination. Nonviable triploid and viable diploid watermelon pollen were removed at similar rates (P = 0.4604). While correlation analyses were not possible for the influence of variable bee abundance on pollen depletion rates, higher bee populations in one year appeared to increase the rate at which pollen grains were removed from staminate flowers.