Older adults tend to rate unfamiliar faces higher on trustworthiness than do their younger counterparts. Although the saying goes “look before you leap”, it is still unknown whether such a strategy could also apply to facial trustworthiness perception, and our understanding of the time course in facial trustworthiness perception also remains unclear. Here, we have argued that a cognitive controlled process suggested by “socioemotional selectivity theory” could potentially lead to such biased trustworthiness perception. Two experiments were conducted to test the association between viewing time and trustworthiness perception. The first study used hierarchical linear modeling in a sample of younger (N = 30, Mage = 20.53, SD = 1.61, 50% female) and older (N = 30, Mage = 63.27, SD = 3.14, 43% female) adults, and found that viewing time and trustworthiness evaluation were positively associated. Using the same stimuli, our second study further manipulated viewing time by two levels (500 ms vs. 3000 ms) and compared younger (N = 28, Mage = 23.93, SD = 2.68, 50% female) and older (N = 30, Mage = 64.47, SD = 4.32, 50% female) adults’ facial trustworthiness evaluation. As expected, a significant three-way interaction revealed that viewing time only impacted older adults’ facial trustworthiness evaluation, and only when given shorter viewing time did older adults show similar facial trustworthiness ratings as younger adults. The present study is the first to directly investigate the relationship between older adults’ viewing time on unfamiliar faces and their perception of facial trustworthiness. Findings suggested that a second thought in facial perception may not benefit older adults’ trustworthiness evaluation.