The objective of the study was to explore the application of a random regression model (RRM) to survival analysis of two dairy breeds: Ayrshire and Jersey. Longevity was defined as the number of months from first calving to culling. The RRM had two fixed factors, province-year of birth and age at first calving. Herd-province-year, additive genetic, permanent environmental and residual effects were included as random factors. Genetic parameters were estimated using Gibbs sampling. At the highest point, heritabilities ranged from 0.08 to 0.21 depending on the breed and model. Genetic parameters were used to estimate breeding values for longevity using BLUP. Results from RRM were compared to EBVs calculated using a multiple trait animal model and a proportional hazards model. RRM for longevity was shown to be feasible and gave rankings that differed from the other two models. More work is needed to determine which of the three models gives the most accurate rankings for longevity.