 Abstract background rush for 54 pyrosequencing has become a popular method for generating transcriptome data from non-model organisms. However, there is no consensus on which software is best suited for assembling these data sets. In this study, we compared five assemblers, CAP3, MIRA, NUBLA, ZECMAN, and CLC, to determine the best practices for transcriptome assembly. Our results showed that NUBLA 2.5 was the most effective assembler, producing longer contigs, better alignment to reference sequences, and being faster and easier to use than the other assemblers. Additionally, combining assemblies from different programs resulted in a more reliable final product. Overall, this study provides valuable insight into the best practices for transcriptome assembly and suggests that combining assemblies from different programs may provide a more accurate and reliable result. This article was authored by Blackster Markel and Kumar Sujai.