The sensitivities of three immunohistological techniques were compared in this study for detecting human immunodeficiency virus (HIV-1) in infected cultured human lymphocytes that had been formalin-fixed and paraffin-embedded. The techniques included in situ hybridization (ISH) with HIV-1 cDNA; immunocytochemistry with HIV-1 p24 monoclonal antibody (ICC-m); and immunocytochemistry with HIV-1 polyclonal antibody from a patient with acquired immunodeficiency syndrome (AIDS) (ICC-p). Procedures were optimized for enzyme digestion and for antibody reaction conditions. HIV-1 -infected cells and noninfected control cells were tested. Noninfected controls were uniformally nega-tive by all three methods. Infected cells had the highest positivity rate by the ISH method (p s O.OOOl), and the ICC-p method was more positive than the ICC-m (p 6 0.0001). Both the ICC-p and the ICC-m techniques were more positive with the cocultivated cell cultures than the ISH, which was more sensitive with the infected continuous cell line (P s 0.0001). The ICC-p method had a lower standard deviation on positive results than either the ICC-m or ISH method. The variability observed with these test procedures, reagents, and specimens suggests that these are important technological parameters in detecting p24, with implications for detecting other HIV-1 markers in infected tissues.