Previous studies have reported two SNPs and a haplotype marker within the Microsomal Transfer Protein gene associated with extreme longevity. Here, we test this finding in a longitudinal study of nonagenarians and in an association study. Participants in the Danish 1905 cohort study (1651 participants aged 92-93 years) were genotyped for the two SNPs (rs2866164 and Q95H) in the Microsomal Transfer Protein gene recently reported to be associated with longevity. The 1905 Cohort has been followed for 6.5 years, during which period 83% of the cohort has died. Furthermore, a group of 575 middle-aged Danish twins (mean age 53.7 years) were tested as a younger control group. The risk haplotype had no significant survival disadvantage (P-values: 0.56, 0.31 and 0.97 in the total population of nonagenarians, males and females, respectively) after 6.5 years of follow-up. The distributions of the suggested risk alleles (rs2866164-G and Q95) and the resulting haplotypes are very similar and not statistically different between the two age cohorts. The frequency for rs2866164-G is in the middle-aged compared to the nonagenarians 25.4 and 23.6% in males and 23.0 and 26.1% in females. The frequency for the risk haplotype is in the middle-aged compared to the nonagenarians 22.7 and 19.2% in males and 18.1 and 21.8% in females. In conclusion, our longitudinal study of survival in the 10th decade of life and an association study in a genetically homogeneous population provided no support for an association between the Microsomal Transfer Protein gene and extreme longevity.