The transcript assembly is essential for transcriptome studies trom next-generation sequencing data. However, there are still many faults of algorithms in the present assemblers, which should be largely improved in the future. According to the requirement of reference genome or not, the transcript assembly could be classified into the genome-guided and de novo methods. The two methods have different algorithms and implementation processes. The quality of assembled transcripts depends on a large number of factors, such as the PCR amplification, sequencing techniques, assembly algorithm and genome character. Here, we reviewed the present tools of transcript assembly and various indexes for assessing the quality of assembled transcripts, which would help biologists to determine which assembler should be used in their studies.