Chromatin immunoprecipitation (ChIP) followed by deep sequencing can now easily be performed across different conditions, time points and even species. However, analyzing such data is not trivial and standard methods are as yet unavailable. Here we present a protocol to systematically compare ChIP-sequencing (ChIP-seq) data across conditions. We first describe technical guidelines for data preprocessing, read mapping, read-density visualization and peak calling. We then describe methods and provide code with specific examples to compare different data sets across species and across conditions, including a threshold-free approach to measure global similarity, a strategy to assess the binary conservation of binding events and measurements for quantitative changes of binding. We discuss how differences in binding can be related to gene functions, gene expression and sequence changes. Once established, this protocol should take about 2 d to complete and be generally applicable to many data sets.