The overall aim of the PhD project was to elucidate the association of human longevity with genetic variation in major candidate genes and pathways of longevity. Based on a thorough literature and database search we chose to apply a pathway approach; to explore variation in genes composing the DNA damage signaling, DNA repair, GH/IGF-1/insulin signaling and pro-/antioxidant pathways. In addition, 16 genes which did not belong to the core of either pathway, however recurrently regarded as candidate genes of longevity (e.g. APOE), were included. In this way a total of 168 genes were selected for investigation. We decided to explore the genetic variation in the form of single nucleotide polymorphisms (SNPs), a highly investigated type of genetic variation. SNPs having potential functional impact (e.g. affecting binding of transcription factors) were identified, so were specific SNPs in the candidate genes previously published to be associated with human longevity. To cover the majority of the common genetic variation in the 168 gene regions (encoding regions plus 5,000 bp upstream and 1,000 downstream) we applied the tagging SNP approach via the HapMap Consortium. Consequently 1,536 SNPs were selected. The majority of the previous publications on genetic variation and human longevity had employed a case-control study design, e.g. comparing centenarians to middle-aged controls. This type of study design is somehow prone to bias introduced by for instance cohort effects, i.e. differences in characteristics of cases and controls, a kind of bias which is avoided when a prospective cohort is under study. Therefore, we chose to investigate 1,200 individuals of the Danish 1905 birth cohort, which have been followed since 1998 when the members were 92-93 years old. The genetic contribution to human longevity has been estimated to be most profound during the late part of life, thus these oldest-old individuals are excellent for investigating such effect. The follow-up survival data enabled performance of longitudinal analysis, which is quite unique in the field of genetic epidemiology of human longevity. Since the cohort was nearly extinct when initiating the PhD study we were able to conduct the longitudinal analyses as regression analyses enabling the estimation of the quantitative effects of the associated SNPs. However, this study explores the genetic contribution to survival during the ninth decade of life, hence, in order to investigate the genetic contribution to survival in younger elderly we also included 800 individuals of the Study of Middle-aged Danish twins (MADT). MADT was initiated in 1998 by random selection of 2,640 intact twin pairs from 22 consecutive birth years (1931-1952) via the Danish Central Person Registry. Only one twin from each twin pair was included in the PhD study. The inclusion of these 800 middle-aged individuals enabled the performance of case-control analysis. Consequently DNA was purified from 2,000 blood samples, genotyping was carried out via the GoldenGate genotyping platform (Illumina Inc.), quality control and data cleaning were conducted, leading to genotype data on 1,394 SNPs in 1,089 oldest-old and 736 middle-aged individuals. The genotype data were analysed at the single-SNP and haplotype levels, and for the 16 candidate genes also at the gene level. The analyses of the data set verified the association of a few of the 16 candidate genes not being part of the candidate pathways under study: SNPs in the APOE, CETP and IL6 genes, while the analyses of the 152 pathway-related genes pointed to new candidate genes of human longevity: especially SNPs in the INS, RAD52 and NTHL1 genes appeared promising. As part of these investigations, replication studies of the single-SNP level findings were conducted in German case-control samples of 1,613 oldest-old (ages 95-110) and 1,104 middle-aged individuals and in a Dutch prospective cohort of 563 oldest-old (age 85+). The Dutch oldest-old have been followed for approximately the same number of years as the Danish oldest-old. Interesting aspects of the study were that the majority of the rare alleles of the identified SNPs were longevity variants, not mortality variants, indicating that at least in our study population, longevity is primarily affected by positively acting minor alleles. Moreover, we observed sex-specific differences in the association of the genetic variation with survival during old age. Furthermore, the genotype data generated were used for a number of replication studies on variation in the FOXO3A, TERT and TERC genes. These studies were performed in response to new data being published on the association of genetic variation in the genes with longevity (FOXO3A and TERT) and with telomere length (TERT and TERC). Our studies verified a role of TERC in human telomere length and of FOXO3A in human longevity (survival from middle age to old age), while a novel role of TERC in human longevity was found. Finally, in addition to the literature and database searches, the genotype data generation and the data analyses mentioned here, RNA purification and qPCR experiments have been initiated in order to investigate gene expression of some of the genes holding SNPs found to be associated with human longevity. Data on one of these genes (IL6) have been included in a manuscript.