This research aimed to experimentally study the anchorage behavior of headed reinforcement in lap splices. Observations of cracking behavior, strain measurements of reinforcement, and strength are reported. The behavior of unconfined laps is compared to confined laps, and the behavior of nonheaded and headed bar laps are also compared. Bar stresses are compared with a proposed model for bearing capacity at the head. Test results suggest that noncontact lap splices should be modeled using a truss mechanism with diagonal compression struts between opposing bars. Under such a model, the capacity of the lap is determined by an anchorage length defined by the intersection points of the diagonal struts. Using this model, the anchorage behavior of headed bars is similar to previously reported results from CCT node tests in which anchorage consisted of bond and head bearing components. Results show that headed reinforcement can markedly reduce required lap length of spliced reinforcement.