This key is designed to be usable by both beginners and non-specialist Hymenopterists alike. It is designed to be the most accurate key available for these groups, regardless of medium of distribution. Each major recent generic key to the Eulophidae, including but not limited to Boucek (1959, 1988), Askew (1968), Schauff (1991), and Schauff, LaSalle, & Coote (1997), has been consulted thoroughly in the preparation of this work, and a serious attempt has been made to critically troubleshoot the potential diagnostic characters of each genus involved. I have attempted to provide the most accurate and informative characters from the available references and from my own studies of specimens. There are still potential problems with this key, but these problems can be fixed with further study.

This key represents an attempt to establish a single reference, as reliable as possible, for the identification of Eulophid genera. I take full responsibility for the information stated in this key, and I take that responsibility seriously. I intend to improve this key on a continuous basis, correcting errors, including new information, and keeping it updated nomenclaturally. I strongly encourage other workers to point out any possible errors, inaccuracies, or omissions discovered in this work, and their efforts will be acknowledged both on the couplet page they improve and in a special acknowledgments page. I would like this page to be a part of a forum for insect identification, absolutely controlled by myself but reflecting the most reliable information available from experts. It can validly be stated that an accurate key to Eulophid genera is impossible to construct given the current state of Eulophid classification. I reply to that by pointing out that this cannot remove the need for serious identification of Eulophids, which this key attempts to serve as well as possible. I certainly recognize that most genera of Eulophidae are imperfectly distinct from related genera, such that an absolutely  accurate key is impossible to make. This key will probably never be perfect, but it is the most ambitious step towards providing an identification reference for Eulophid genera, and it is intended to improve as the classification of the family reaches a more satisfactory state. The alternative is to rely upon highly unreliable references that frustrate rather than aid the audience that they are intended to serve. This work does not provide a perfect alternative, but is intended to represent an improvement over previously available identification references. It is not intended to aid experts who already have a dynamic and highly accurate notion of the genera keyed here, as they do not and should not need keys for their group of specialty, but their aid in improving this key will be important in determining the rate of its improvement. I am aware of some areas in this key that should prove difficult to navigate, and they are discussed in a special page entitled, Problems and resources needed.

Current valid names under the ICZN code of nomenclature will be followed in this work, despite any disagreement that has not resulted in formal nomenclatural changes. In the event that internet publication of nomenclatural changes actually become valid or common, this page will never include attempted nomenclatural changes, and is not intended to establish such changes in any way. Unpublished names or nomenclatural changes will not be posted in this site until the work in which they are published is printed and distributed.

This key is intended to be useful only in identification, and not as a strong source of phylogenetic information, although some speculation concerning phylogenetic relationships is included in comments about certain genera. The arrangement of couplets in this key is intended to reflect overall similarity and difficulty in distinguishing included genera, and has never and will never be altered reflect phylogenetic information at the expense of accurate identification. The similarity of dichotomous keys to dendrograms has misled many  workers to produce keys that reflect preferred phylogenetic hypotheses, but have little or no utility for comparing organisms that are easily confused in practical attempts at identification. This mistake will be avoided here as much as possible.

There has been much recent furvor (mostly in personal communication) about wording in key couplets. I have made a special effort to avoid using vague or misleading terminology and logical formulae in my key. Some uses of the word "usually" cannot be avoided due to the lack of morphological uniformity in Eulophid genera, but use of this word has been kept to the minimum possible without sacrificing accuracy, and in all cases some universal character has been used as the primary character backed up by the "usually" clauses. Frequently, the universal character is one that most researchers avoid examining, such as a genitalic or sensillar character, and the use of non-universal characters at that point seems advisable. Simple logical formulae, such as if...then, and..., and/or, etc. statements have been used when necessary, but in many cases these have been avoided by making the genus key to more than one terminus. Whenever used, the words constructing the formulae are bolded. Also, whenever genera key to more than one terminus, the omission of qualifiers denoting special exceptional groups of a genus (such as species or species groups that violate the conventional definition of a genus) has proven frustrating in my efforts to use keys in general. It is impossible in those cases for the user to know if only one rare species of a genus keys to that terminus or if many common species key there. I have made a special effort, in the cases where a genus keys to more than one terminus, to note if most species, only rare forms, or only certain species groups key to a particular terminus.

This is entirely an internet-based work, and it is designed to reflect the obvious advantages of internet media over bound printed media. This is not to say that bound printed media do not have advantages as well, and I fully recognize that many or most workers will greatly prefer the printed version. The couplets comprising this key can be printed individually, but the included images will likely be of very poor quality. This is a problem that I hope to remedy in the future. Furthermore, it will probably be noticed by visitors to this site that the pages I design are simple in comparison to most other web pages. This is not because I am unskilled in making web pages, but because I prefer simplicity in identification references and broad compatibility in web pages designed to be useful to workers using any platform or computer system. The superfluous effect provided by entertaining decorations is deliberately sacrificed so that I have more time to construct keys that are as useful and accurate as possible.

Finally, I would like to state that it is my goal to make many more of these keys. For most groups, this requires an extensive amount of background research. The literature alone is of course not sufficient for constructing a key to anything, let alone a key aspiring to be as good as I want mine to be. With further study, and with visits to large entomological collections, I hope to construct presentable keys to genera of other Hymenopteran groups, such as Pteromalidae, Encyrtidae, Platygastrinae, Megaspilidae, and so on. Also, I believe that my system is especially well-suited for keys to species. I hope to demonstrate this in the coming years. I do not have a problem collaborating with individuals who would like to make use of my method, and I will devote as much effort as possible to helping bring about keys that use my method. Collaboration would certainly make my life easier, as a person's lifetime is limited and I can't learn every group I would like to make keys for. I would like to caution, however, that I think it best if I have extensive input into how the key is made. I cannot/will not prevent people from using my method, but I wish to be active in improving the science of diagnostic tool construction, and I would like to do as much as I can to further this goal. Also, it would be preferable to have a standard source of keys that can be trusted, and I am driven to bring about such a source.