While many cell components can be studied by pre- or postembedment immunohistology, one would like to possess an alternative method in order to overcome all problems observed in pre- and postembedment staining due to penetration difficulties e.g. for antibodies and cell alterations by dehydration and resin curing. Thus, the development of cryoultramicrotomy offered great promise for immunostaining of cells to be done on ultrathin frozen sections without classical procedures of dehydration and plastic embedment. Several immunohistological procedures including various marker systems were tried with ultrathin frozen sections, but until now reasonable results were obtained only with ferritin as label. Hitherto, direct or indirect peroxidase stainings were most often disappointing inasmuch as the original 3,3′-diamino-benzidine method was very capricious when applied to ultrathin frozen sections. In this note we describe our recent progress. We have used simian adenovirus type 7 (SA7) infected monkey kidney cells in order to localize virus antigens in the cell nucleus. We are aware that the employed anti-human adeno 7 immune sera possessed antibody specificities against the bulk of virus antigens, but these immune sera are sufficient to demonstrate feasibility of the immunoperoxidase method in cryoultramycrotomy.