Large numbers of new taxa are described annually and while there is a great need to make them identifiable, there seems little consistency in how this might be facilitated. I surveyed 427 papers published in 2021 and 2022, which described 587 new insect genera. I found that only 136 of the papers included keys, and these allowed the identification of 233 of the new genera 31.9% of papers and 39.7% of the new genera). The proportion of papers that included a key varied significantly among insect orders but not among the handful of journals wherein the bulk of the new genera were described. Overall, for 17 key-related variables assessed in a binary fashion (optimal vs suboptimal), the average key had almost six criteria that were scored as being suboptimal. For example, less than one-fifth facilitated retracing and less than 12% had illustrated keys where the images were conveniently located close to the relevant key couplets. Progress towards confirming a putative identification was possible in all papers, through the inclusion of a diagnosis, habitus images or both.

Based upon this analysis, and expanding on previous suggestions for key construction, I make 23 recommendations on how to make an identification key maximally useful for users and indicate the relative ease with which each could be adhered to. I suggest that identification keys should accompany all new taxon descriptions, that guidelines for effective key construction be added to journals’ instructions to authors, that editors and reviewers should check keys carefully and that publishers should be attentive to the needs of users through, for example, permitting duplication of images to make keys easier to use. My recommendations are likely relevant to all levels in the taxonomic hierarchy for all organisms, despite the data being derived from generic-level keys for insects.