Triangulation is a research strategy that has received considerable practical and theoretical interest in nursing research. However, despite its popularity, triangulation raises a number of problematic methodological and philosophical issues. Four such issues are discussed in this paper. First, it is argued that triangulation is not always a necessary ingredient of a research project, as some questions, framed in fairly specific terms, may be adequately answered with the use of a single method. Second, triangulation is often advanced as a means of securing the validity of data; however, this claim often makes unwarranted assumptions as to the criterial status of one method in relation to others. To the extent that triangulation does secure validity, this is likely to be in respect of content validity (i.e. the scope of findings) rather than, as is often suggested, criterion-related validity (i.e. the 'accuracy' of findings). Third, the bringing together of quantitative and qualitative methods in triangulation tends to ignore fundamental questions as to whether data generated by methods that come from divergent epistemological frameworks can meaningfully be combined. Finally, the use of different theoretical frameworks, in theory triangulation, mistakes the true nature of a theory and overlooks the distinctness and incommensurability of different theoretical perspectives. In conclusion, nurse researchers should approach triangulation in a cautious and critical way, with due regard to the rather dubious assumptions upon which it frequently rests.