Crossing structures, commonly known as bridges, are used by conservationists to connect fragmented wildlife habitat over roads. Crossing structures show varying levels of success. One factor that may explain success, or lack thereof, is crossing structure design. We assessed three different bridge models in a semi-experimental setting. Materials, costs, and design complexities were relatively similar among the models. Two New World monkey species (Alouatta palliata, n=13; Ateles geoffroyi, n=2) in a Costa Rican wildlife sanctuary were given uniform access to the models: twisted liana, parallel lianas, and horizontal ladder. Model performance was evaluated by observing study subjects' behavior during 38 sampling hours in July 2013. In general, crossing structures were rarely used during the study period, comprising only 0.1% of all scan samples. Although adult study subjects represented 47% of the study group, adults used the crossing structures most often (89% of all cases). Overall, there was no difference in frequency of use among models (X2= 0.89, df = 2, p = 0.64). However, in 48% of cases (n=13) study subjects exhibited difficulty traveling across the crossing structures or showed signs of fear that were attributable to the crossing structures. Of this subset, the majority of problems (62%) were associated with the twisted liana model. Preliminary findings show that the ladder and parallel lianas performed better than the twisted liana, but our sample size is small and thus continued investigation is necessary for validation. For future experiments we recommend additional research on improving structural stability and incorporating living vegetation in the designs.