This work discusses the criteria proposed to consider wild bees as bioindicators and specifically applied to orchid bees in neotropical forests. Some of the issues are: 1) the deficiencies of the sampling methods, which makes it difficult to accurately assess species inventories. 2) missing knowledge about the biology of many species. 3) spatial or temporal distribution of most species remains unknown, which may misslead the results of short-term studies. 4) It is not clear whether orchid bees are affected by climate change as seen in other bees, which weakens their predictive potential. 5) A measure of the economic benefits provided by orchid bees is needed to better appraise them and their conservation. Finally, future studies should develop predictive models for conservation, accounting for evolutionary aspects like phylogeny or distributions; together with studies of the effect of disturbance on the physiology of the bees.