We examined the effects of repeated inbreeding on fitness components of the long-lived perennial Succisa pratensis (Dipsacaceae). Plants from six populations differing in size were used to establish lines with expected inbreeding coefficients f of 0, 0.5 and 0.75. The effects of different inbreeding levels were measured for seed set, seed mass, percentage germination and seedling relative growth rate. Seed set decreased following one generation of inbreeding and seedling growth rate decreased after two generations of inbreeding. Our study indicated that the mutational load is difficult to purge and that continued inbreeding tends to affect important traits in S. pratensis. Although the partial dominance hypothesis for inbreeding depression seems to account for the results, the overdominance hypothesis cannot be ruled out completely. Overall, we conclude that the response of a long-lived plant, such as S. pratensis, to repeated inbreeding does not differ from that of other plant species with shorter life spans, surely because the mechanisms that account for inbreeding depression are universal for all plant species.