Despite the strong evidence for a major role played by genetic factors in the aetiology of non-insulin-dependent diabetes mellitus (NIDDM), the genes involved are still unknown. Association studies of candidate genes for the inheritance of NIDDM have so far yielded inconclusive results. Some evidence exists for an association between NIDDM and the glucose transporter gene GLUT1, involved in basal glucose transport, although this has not been confirmed. In the present study we have tested the hypothesis of linkage between NIDDM and the GLUT1 gene, using affected sib-pairs. With this method the concordance observed for a given gene marker is compared with that expected under the assumption of no linkage between that marker and the disease. Fifty-four pedigrees (22 Italians and 32 British), for a total of 82 sib-pairs were studied by the affected sib-pair method proposed by Weeks and Lange, using two restriction fragment length polymorphisms (RFLPs) at the GLUT1 locus, the MspI RFLP, at an estimated 0.171 recombination frequency from the GLUT1 gene, and the XbaI RFLP, located within the GLUT1 gene and previously shown to be associated with the disease. Results showed that the MspI marker and NIDDM segregate independently; for the XbaI RFLP, linkage could be shown only if the results were weighted by the allele frequency [f(p) = 1/p], and only in the Italian and the combined (Italian and British) sib-pair groups. Multilocus analysis with both markers was also negative. We conclude that the GLUT1 gene is very unlikely to play a major role in the aetiology of NIDDM, although an accessory role cannot be excluded, and studies of the gene sequence should help to clarify this question.