Most studies on lung function heritability have been conducted in smokers and non-smokers using cross-sectional study design. Smoking patterns may, however, confound the contribution of genetic factors. We investigated heritability of forced expiratory volume in one second (FEV1), forced vital capacity (FVC), and FEV1/FVC ratio longitudinally, excluding the effects of smoking. A sample of never smoking female twins (n = 374), aged 63-76 at baseline, answered health questionnaires and attended spirometry in years 2000 and 2003. Bivariate structural equation modeling, restricted to adequate spirometry performances (baseline n = 339, follow-up n = 252), was used to estimate genetic and environmental influences on consecutive measurements of FEV1, FVC, and FEV1/FVC. The best-fitting models included additive genetic and non-shared environmental effects. Heritability estimates of 32% and 36% for FEV1, 41% and 37% for FVC, while 46% and 16% for FEV1/FVC were found at baseline and at follow-up. Genetic correlation between FEV1 and FEV1/FVC heritability estimates approached unity, whereas correlation between FVC estimates was 0.80. Environmental correlations were 0.69 for FEV1, 0.62 for FVC, and 0.07 for FEV1/FVC. In never smokers, additive genetic and non-shared environmental effects explain the inter-individual variations in FEV1, FVC, and FEV1/FVC. One third of the variation in FEV1 and FVC is explained by genetic and two thirds by environmental effects. Between 2000 and 2003, environmental effects on FEV1/FVC changed, and the proportion of variance explained by environmental effects increased remarkably. Genetic effects on FEV1 and FEV1/FVC are common to consecutive measurements, whereas at follow-up, new genetic factors explained 14% of the observed variance in FVC.