Dispersal is a biological process performed in three stages: emigration, transfer and immigration. Intra-specific variation on dispersal behavior, such as sex-bias, is very common in nature, particularly in birds and mammals. However, dispersal is difficult to measure in the field and many hypotheses concerning the causes of sex-biased dispersal remain without empirical confirmation. An important limitation of most empirical studies is that inferences about sex-biased dispersal are based only on emigration proneness or immigration success data. Thus, we still do not know whether sex-biased immigration in fragmented landscapes occurs during emigration, transfer or in both stages. We conducted translocation and radiotracking experiments to assess i) whether inter-patch dispersal movements of a rainforest bird (Pyriglena leucoptera) is sex-biased and ii) how dispersal stages and the perceptual range of the individuals are integrated to generate dispersal patterns. Our results showed that inter-patch dispersal is sex-biased at all stages for P. leucoptera, as females not only exhibit a higher emigration propensity but are subjected to a lower risk of predation when moving through the matrix. Moreover, our data support a perceptual range of 80 m and our results showed that dispersal success decreases considerably when inter-patch distances exceeds this perceptual range. In this case, birds have a higher probability of travelling over longer routes and, as a consequence, the risk of predation increases, specially for males. Overall, results supported that assuming dispersal as a single-stage process to describe dispersal behavior may be misleading. In this way, our study advanced our understanding of processes and patterns related to inter-patch dispersal of neotropical forest birds, shedding light on potential implications for population dynamics and for the management of fragmented landscapes.