In clinical practice, parasomnias are often found to run in families and to co-occur. Several studies have indicated a role of genetic factors in them. In 1990, a questionnaire (response rate, 77%) sent to the Finnish Twin Cohort, a representative population sample aged 33-60 years, surveyed the frequency of five parasomnias (sleepwalking, sleeptalking, enuresis, bruxism, and nightmares) in childhood and as adults. In assessing the phenotypic covariation and shared genetic effects between the parasomnias, we used polychoric correlations and structural equation modelling. In childhood (n = 5856 individuals), co-occurrence is highest in sleeptalking with sleepwalking (R = 0.73), nightmares (R = 0.50), and bruxism (R = 0.43). As adults (n = 8567), the results are similar (R = 0.56, 0.43, and 0.39, respectively). The analyses of shared genetic effects included 815 monozygotic and 1442 dizygotic twin pairs with complete responses on four parasomnias as adults. The strongest genetic covariation was found in sleeptalking with sleepwalking, sleeptalking with bruxism, and in sleeptalking with nightmares. The estimated proportions of shared genetic effects were 50, 30, and 26%, respectively. The present results indicate that parasomnias share some common genetic background.