Hubbard breeder pullets were fed a complete pullet developer ration on an ad libitum (AL) or restricted feeding (RF) regimen. The ration either was unmedicated or contained .0125% amprolium, .0125% clopidol, or .0110% monensin. The relationship between the feeding schedule and coccidial infection was determined on the basis of 1) efficacy of the medication in controlling a single infection in susceptible pullets and 2) the development of immunity to subsequent challenge inoculation after a series of immunizing inoculations. Cage-reared, susceptible pullets were inoculated with sporulated oocysts of either. Eimeria tenella, 2 strains of E. acervulina, or E. maxima. With all three medications, the infection with a least one species was more severe, as measured by intestinal lesion score, in the RF pullets than in the corresponding AL pullets. Other pullets were kept for three weeks in floor pens that contained coccidial oocysts to allow natural infection and immunity to develop. The pullets were then transferred to suspended cages to prevent reinfection and fed an unmedicated ration. After one week, the pullets were challenged with the same coccidial strain as that used for immunizing. All pullets initially exposed to coccidia and given no medication were resistant to challenge inoculation. Control pullets are exposed to coccidia during rearing were susceptible to challenge. The administration of anticoccidial drugs (especially monensin) by ad libitum feeding interfered with development of immunity under these conditions, but, when the same drugs were given with a restricted feeding regimen, they did not interfere with development of immunity.