We developed a standardized protocol for comparing the sensitivities of the embryos of different bird species to methylmercury when methylmercury was injected into their eggs. During the course of developing this protocol, we investigated the effects of various factors on the toxicity of the injected methylmercury. Most of our experiments were done with chicken (Gallus domesticus), mallard (Anas platyrhynchos), and ring-necked pheasant (Phasianus colchicus) eggs, all of which were purchased in large numbers from game farms. A smaller amount of work was done with double-crested cormorant (Phalacrocorax auritus) eggs collected from the wild. Several solvents were tested, and corn oil at a rate of 1 :l/g egg contents was selected for the final standardized protocol because it had minimal toxicity to embryos and because methylmercury dissolved in corn oil yielded a dose?response curve in a range of egg concentrations that was similar to the range that causes reproductive impairment when the mother deposits methylmercury into her own eggs. The embryonic stage at which eggs were injected with corn oil altered mercury toxicity; at early stages, the corn oil itself was toxic. Therefore, in the final protocol we standardized the time of injection to occur when each species reached the morphologic equivalent of a 3-day-old chicken embryo. Although solvents can be injected directly into the albumen of an egg, high embryo mortality can occur in the solvent controls because of the formation of air bubbles in the albumen. Our final protocol used corn oil injections into the air cell, which are easier and safer than albumen injections. Most of the methylmercury, when dissolved in corn oil, injected into the air cell passes through the inner shell membrane and into the egg albumen. Most commercial incubators incubate eggs in trays with the air cell end of the egg pointing upward, but we discovered that mercury-induced mortality was too great when eggs were held in this orientation. In addition, some species of bird eggs require incubation on their sides with the eggs being rolled 180? for them to develop normally. Therefore, we adopted a procedure of incubating the eggs of all species on their sides and rolling them 180? every hour. Little has been published about the conditions of temperature, humidity, and the movements to which eggs of wild birds need to be subjected for them to hatch optimally under artificial incubation. Not unexpectedly, hatching success in an artificial incubator is generally less than what natural incubation by the parents can achieve. However, the survival of control embryos of most wild bird species was good (generally > 80%) up to within 1 or 2 days of hatching when we incubated the eggs at 37.5?C (or 37.6?C for gallinaceous species) at a relative humidity that resulted in an approximate 15% to 16% loss in egg weight by the end of incubation and by incubating the eggs on their sides and rolling them 180?/h. To improve statistical comparisons, we used survival through 90% of incubation as our measurement to compare survival of controls with survival of eggs injected with graded concentrations of mercury.