For quantitative real-time PCR (qRT-PCR) analysis of gene expression in plants,the key prerequisite that determines the accuracy of results is the selection of appropriate reference gene(s).Goji (Lycium barbarum) is a multi-branched shrub belonging to the Solanaceae family.Despite its outstanding stress resistance and economic value,no systematic screening or evaluation of reference gene(s) in L.barbarum has been performed,mainly due to the lack of its whole genome data.In this work,we identified 18 candidate reference genes from the transcriptomic sequencing data of 14 samples obtained from the leaves,flowers,and fruits of L.barbarum at different developmental stages and under drought stress condition.The expression stability of these candidate genes was rigorously analyzed using qRT-PCR and four different statistical algorithms: geNorm,BestKeeper,NormFinder and RefFinder.We found that two novel reference genes LbCML38 and LbRH52 showed the most stable expression.On the contrary,traditionally used reference genes such as LbGAPDH,LbHSP90 and LbTUB showed unstable expression in the tested samples.