In collaborative qualitative research in Asia, data are usually collected in the national language, and this poses challenges for analysis. Translation of transcripts to a language common to the whole research team is time consuming and expensive; meaning can easily be lost in translation; and validity of the data may be compromised in this process. We draw on several published examples from public health research conducted in mainland China, to highlight how language can influence rigour in the qualitative research process; for each problem we suggest potential solutions based on the methods used in one of our research projects in China. Problems we have encountered include obtaining sufficient depth and detail in qualitative data; deciding on language for data collection; managing data collected in Mandarin; and the influence of language on interpreting meaning. We have suggested methods for overcoming problems associated with collecting, analysing, and interpreting qualitative data in a local language, that we think help maintain analytical openness in collaborative qualitative research. We developed these methods specifically in research conducted in Mandarin in mainland China; but they need further testing in other countries with data collected in other languages. Examples from other researchers are needed.