Premise: Flowering plants with poricidal anthers are commonly visited by buzzing bees, which vibrate flowers to extract pollen. However, not all flower visitors are in fact pollinators, and features such as body size and duration of flower visits are important factors in determining pollination effectiveness. We tested whether bee-to-flower size relationships predict the pollination effectiveness of flower visitors of a buzz-pollinated species (Chamaecrista ramosa, Fabaceae). Methods: We sorted 13 bee taxa into three groups: smaller than, equivalent to (“fit-size”), and larger than flower herkogamy (spatial separation between anthers and stigma). We expected the latter two groups to touch the stigmas, which would be an indicator of pollination effectiveness, more frequently than the first group. To test this hypothesis, we assessed contact with stigmas, foraging behavior, and duration of visits for the three size groups of bees. Results: Our data reveal that small bees scarcely touched the stigmas, while large and fit-size bees were the most efficient pollinators, achieving high stigma-touching rates, conducting much shorter flower visits, and visiting flowers and conspecific plants at high rates during foraging bouts. Conclusions: The results did not show size-matching among bees and flowers, as expected, but rather a minimum size threshold of efficient pollinators. The finding of such a threshold is a nonarbitrary approach to predicting pollination effectiveness of visitors to herkogamous flowers with poricidal anthers.