The initiation of T-cell responses requires rare precursors to locate a draining lymph node (dLN) and encounter dendritic cells (DCs) presenting peptide-major histocompatibility complexes (pMHCs). To locate this needle in the haystack rapidly, T cells face an optimization problem-what is the most efficient trafficking strategy for surveillance and recirculation through blood? Two extremes are scanning low numbers of DCs per node with frequent recirculation, or meticulous surveillance with infrequent recirculation. Naive T cells also require stimulation by self-pMHCs. To enable efficient location of both foreign and self, has evolution settled on an optimum time for T cells to spend surveying each lymph node? Using a data-driven mathematical model, we show the most efficient strategy for detecting antigen in a dLN depends on its abundance. Detection of low-density antigen is optimized with systemically slow transit. In contrast, at high densities or if dLN egress is restricted, rapid transit through other nodes is optimal. We argue that blood-lymph recirculation dynamics facilitate a trade-off, and are consistent with dominant roles for the very early detection of rare foreign antigens in a dLN, and the efficient accumulation of signals from systemically distributed self-antigens.