The generation of transgenic mice by microinjection of DNA into the pronuclei of fertilized oocytes was described in the early 1980s. A number of parameters affecting the efficiency of the technique were soon identified, including the type of DNA construct, the concentration of DNA being injected, and, most importantly, the strain of mice used for oocyte donors. Since then, hundreds of laboratories and transgenic core facilities across the world have successfully used this technique, essentially as originally described, to create thousands of new transgenic mouse lines. However, the overall procedure continues to be relatively inefficient, in terms of the number of fertilized oocytes required to produce a transgenic mouse, and variations in yields from day to day and construct to construct can be large. Consequently, core facilities often struggle to explain to their customers why a sufficient number of transgenic founders were not produced from a given construct. We believe the field (and individual facilities) would benefit from a rigorous assessment of average yields and expected variations in yields. To this end, we have initiated a survey from the International Society for Transgenic Technologies (ISTT) web site ( www.transtechsociety.org ), to obtain raw microinjection data from as many facilities as possible. We intend to use this data to establish performance standards for the field. Existing facilities will be able to refer to these standards in dealing with dissatisfied clients, and new facilities will be able to aim for an achievable goal. We may even be able to discover an optimum combination of factors that will allow every facility to achieve higher yields.