The evidence that there is clinical heterogeneity of type 1 diabetes is reviewed and the implications for genetic studies are discussed. In the past year, genome-wide linkage analysis of 1435 multiplex families was reported. Additionally, confirmed evidence for association of specific markers at two loci (PTPN22, OAS1) as well as failure to replicate three others (IL12B, SUMO4, PAX4) is discussed. Some common themes are identified and suggestions for improvements are made. We look forward to the results from genome-wide association studies.