The steps of two immunofluorescent-antibody-based detection methods were evaluated for their efficiencies in detecting Giardia cysts and Cryptosporidium oocysts. The two methods evaluated were the American Society for Testing and Materials proposed test method for Giardia cysts and Cryptosporidium oocysts in low-turbidity water and a procedure employing sampling by membrane filtration, Percoll-Percoll step gradient, and immunofluorescent staining. The membrane filter sampling method was characterized by higher recovery rates in all three types of waters tested: raw surface water, partially treated water from a flocculation basin, and filtered water. Cyst and oocyst recovery efficiencies decreased with increasing water turbidity regardless of the method used. Recoveries of seeded Giardia cysts exceeded those of Cryptosporidium oocysts in all types of water sampled. The sampling step in both methods resulted in the highest loss of seeded cysts and oocysts. Furthermore, much higher recovery efficiencies were obtained when the flotation step was avoided. The membrane filter method, using smaller tubes for flotation, was less time-consuming and cheaper. A serious disadvantage of this method was the lack of confirmation of presumptive cysts and oocysts, leaving the potential for false-positive Giardia and Cryptosporidium counts when cross-reacting algae are present in water samples.