The large size of conventional antibodies impedes tissue penetration and renal elimination, resulting in suboptimal in vivo targeting. Here we assess the utility of nanobodies and nanobody-Fc-fusion proteins as alternatives to monoclonal antibodies as theranostics, using T cell ADP-ribosyltransferase 2 (ART2) as a model antigen for specific targeting of lymph nodes. ART2-specific monovalent nanobody s + 16a (17 kDa), a bivalent Fc-fusion protein of s + 16a (s + 16-mFc, 82 kDa), and conventional antibody Nika102 (150 kDa) were labeled with AlexaFluor680. In vitro binding and inhibitory properties of the three AF680 conjugates were assessed by flow cytometry. For in vivo imaging experiments, AF680 conjugates were intravenously injected in mice lacking (KO) or overexpressing (TG) ART2. We monitored circulating and excreted AF680 conjugates in plasma and urine and performed in vivo near-infrared fluorescence imaging. Nanobody s + 16a(680) and s + 16mFc(680) labeled and inhibited ART2 on T cells in lymph nodes within 10 min. In contrast, mAb Nika102(680) required 2 h for maximal labeling without inhibition of ART2. In vivo imaging revealed specific labeling of ART2-positive lymph nodes but not of ART2-negative lymph nodes with all AF680 conjugates. Even though bivalent s + 16mFc(680) showed the highest labeling efficiency in vitro, the best lymph node imaging in vivo was achieved with monovalent nanobody s + 16a(680) , since renal elimination of unbound s + 16a(680) significantly reduced background signals. Our results indicate that small single-domain nanobodies are best suited for short-term uses, such as noninvasive imaging, whereas larger nanobody-Fc-fusion proteins are better suited for long-term uses, such as therapy of inflammation and tumors. Copyright © 2014 John Wiley & Sons, Ltd.