The nonlinear response of microbubbles to ultrasound is useful for high-sensitivity, real-time imaging of intravascular structures. Molecular imaging is achieved by adhering the microbubbles to sites on diseased endothelium through specific ligand-receptor interactions. Conventionally, the microbubble surface architecture is engineered for maximal exposure of the targeting ligand to the surrounding milieu, in order to increase the probability that the ligand will find its receptor. Ligand presentation is a double-edged sword, however, because its exposure to blood can induce an immune response. Complement activation, which is a cascade of events amplified at each step, is of particular concern. A new targeting strategy is presented that is based on ultrasonic manipulation of ligand presentation within the ultrasound beam focus. The microbubble surface architecture is designed to be stealth in the unperturbed state – i.e., the tethered ligand is buried by a polymeric overbrush that conceals it from plasma components. Application of ultrasound radiation force not only pushes the microbubble against the endothelium, but also transiently reveals the ligand for specific adhesion. This affords remote operator control over both physical and chemical targeting. The current work focuses on theoretical considerations and new immunogenicity and adhesion experiments. Model calculations show how entropy-driven vertical segregation in the bimodal brush leads to ligand concealment from plasma components and how gas-core oscillations in a low-amplitude ultrasound field transiently reveal the ligand. Immunogenicity experiments in vitro and in vivo show significantly reduced complement activation by the buried-ligand architecture. Five minutes after injection of microbubbles into mice, for example, the level of serum C3a anaphylatoxin rose by 2.2-fold for the exposed-biotin, but only 1.5-fold for the buried-biotin, versus no-biotin control (p<0.01). Moreover, after a 10-minute incubation of microbubbles in human serum, the amount of C3/C3b bound to microbubbles increased by 20-fold and the amount of soluble C3a increased by 5-fold compared to shielded-biotin or no-biotin control architectures (p<0.01). This corroborates and explains earlier results (Molecular Imaging, 2006, 5:139), which showed that burying the ligand enhanced contrast persistence in vivo without loss of specific adhesion in vitro mediated by ultrasound radiation force. Discussion will also include recent in vitro adhesion results for physiological peptide ligands.