Model-based analysis of ChIP-seq (MACS) is a computational algorithm that identifies genome-wide locations of transcription/chromatin factor binding or histone modification from ChIP-seq data. MACS consists of four steps: removing redundant reads, adjusting read position, calculating peak enrichment and estimating the empirical false discovery rate (FDR). In this protocol, we provide a detailed demonstration of how to install MACS and how to use it to analyze three common types of ChIP-seq data sets with different characteristics: the sequence-specific transcription factor FoxA1, the histone modification mark H3K4me3 with sharp enrichment and the H3K36me3 mark with broad enrichment. We also explain how to interpret and visualize the results of MACS analyses. The algorithm requires ∼3 GB of RAM and 1.5 h of computing time to analyze a ChIP-seq data set containing 30 million reads, an estimate that increases with sequence coverage. MACS is open source and is available from http://liulab.dfci.harvard.edu/MACS/.