Patterns of host-parasite association are poorly understood in tropical forests. While we typically observe only snapshots of the diverse assemblages and interactions under variable conditions, there is a desire to make inferences about prevalence and host-specificity patterns. We studied the interaction of ticks with non-volant small mammals in forests of Borneo. We inferred the probability of species interactions from individual-level data in a multi-level Bayesian model that incorporated environmental covariates and advanced estimates for rarely observed species through model averaging. We estimated the likelihood of observing particular interaction frequencies under field conditions and a scenario of exhaustive sampling and examined the consequences for inferring host specificity. We recorded a total of 13 different tick species belonging to the five genera Amblyomma, Dermacentor, Haemaphysalis, Ixodes, and Rhipicephalus from a total of 37 different host species (Rodentia, Scandentia, Carnivora, Soricidae) on 237 out of 1,444 host individuals. Infestation probabilities revealed most variation across host species but less variation across tick species with three common rat and two tree shrew species being most heavily infested. Host species identity explained ca. 75 % of the variation in infestation probability and another 8-10 % was explained by local host abundance. Host traits and site-specific attributes had little explanatory power. Host specificity was estimated to be similarly low for all tick species, which were all likely to infest 34-37 host species if exhaustively sampled. By taking into consideration the hierarchical organization of individual interactions that may take place under variable conditions and that shape host-parasite networks, we can discern uncertainty and sampling bias from true interaction frequencies, whereas network attributes derived from observed values may lead to highly misleading results. Multi-level approaches may help to move this field towards inferential approaches for understanding mechanisms that shape the strength and dynamics in ecological networks.