Ligation-based methods and kits are disclosed for identifying at least two target nucleotides in a mixed population sample, that is a sample that contains or potentially contains target nucleic acid sequences from more than one source. Typically, two ligation reaction compositions are formed, ligation products generated, and the ligation products or their surrogates are analyzed to identify target nucleotides in the mixed population sample. In certain embodiments, the target nucleic acid sequences, the ligation products, or both are amplified. In certain embodiments, multiplex amplification and/or ligation reactions are performed.