Chromatin immunoprecipitation followed by deep sequencing (ChIP-seq) offers a powerful means to study transcription factor binding on a genome-wide scale. While a number of advanced software packages have already become available for identifying ChIP-seq-binding sites, it has become evident that the choice of the package together with its adjustable parameters can considerably affect the biological conclusions made from the data. Therefore, to aid these choices, we have recently introduced a reproducibility-optimization procedure, which computationally adjusts the parameters of the popular peak detection algorithms for each ChIP-seq data separately. Here, we provide a detailed description of the procedure together with practical guidelines on how to apply its implementation, the peakROTS R-package, in a given ChIP-seq experiment.