A common measure of anti-predator defense in birds under field conditions is the escape-related distances, which an individual would take action (e.g. alert, ?ee) when approached by a potential predator (e.g. human). These distances represent a compromise between the risks of mortality due to predation and foraging opportunities. The difference of escape-related distances among groups should reflect the difference in expected probability of survival: the group with relatively good survival prospect can be predicted to take small risks (long escape-related distances) in order not to jeopardize their prospects of survival. In addition, urban birds should have shorter escape-related distances in order to coexist with humans that caused frequent disturbance than rural ones, while larger flocks have more eyes that should result in an earlier detection of approaching predators, therefore, increasing escape-related distances. We used field data on age-specific escape-related distances in Tree Sparrows Passer montanus (145 adults and 75 juveniles) from Beijing city to test these predictions (Fig. 1 & 2). The analyses showed longer escape-related distances in adult than juvenile (Figs. 3 & 4) (e.g. alert distance in adults 11.76 ± 0.60 m, in juveniles 9.33 ± 0.71 m, Mann-Whitney U test, Z = 2.662，P = 0.008; flight initiation distance in adults 8.71 ± 0.53 m, in juveniles 7.31 ± 0.68 m, Mann-Whitney U test, Z = 1.872，P = 0.061), and the adults, rather than juveniles, have higher probability to be the first individuals to escape within the flock (χ2 = 4.934，df = 1，P = 0.026). Independently, escape-related distances were significantly negatively related with the number of pedestrians, and positively correlated with flock size (Table 1). These findings are consistent with the hypothesis that adults take smaller risks than juveniles.