The behavioral study of elusive species, which often cannot be directly observed in the wild, is based on the use of animal monitoring technologies. Here, we evaluate the performance of four methods applied simultaneously on a single population to detect and characterize the activity behavior of elusive species, using the giant armadillo Priodontes maximus as a model. GPS tagging, accelerometer, and camera trap array data were used to characterize general activity patterns, while camera traps placed in front of burrows in use were used to define activity duration, onset, and termination times. Circular kernel probability density functions were used to characterize activity patterns with each data set and the overlap coefficient D4 was used to characterize the similarity between them. We gathered data from 53 armadillos during 9 years in the Pantanal wetlands of Brazil. Activity duration estimates showed significant variation among monitoring methods but were generally short (4.5-6 h on average). Overall, the estimates of activity timing provided by the methods used were relatively similar (>75% D4 overlap) and pointed to a nocturnal activity pattern. Armadillos left their burrows mainly in the first hours after sunset, and terminated their activity mainly between midnight and 4 AM. However, the camera trap array provided a later activity onset and peak estimates for the species when compared to other methods. The species spends 75-80% of its time underground and accelerometer data shows that armadillos are not active and do not feed while underground, pointing to a fossorial behavior for the species. Although researchers should be aware of the small biases that can emerge from some of the methods, the general similarities among patterns indicates that researchers can choose among the ever-increasing number of animal monitoring technologies depending on their research question and constraints.