Despite its potential implications for the objectivity of scientific knowledge, the claim that ‘scientific instruments are perspectival’ has received little critical attention. I show that this claim is best understood as highlighting the dependence of instruments on different perspectives. When closely analysed, instead of constituting a novel epistemic challenge, this dependence can be exploited to mount novel strategies for resolving two old epistemic problems: conceptual relativism and theory ladeness. The novel content of this paper consists in articulating and developing these strategies by introducing two fine-grained notions of perspectives as the key units of analysis: ‘broad perspectives’ and ‘narrow perspectives’