A dynamic and stochastic simulation model (the "JohneSSim model") was developed to evaluate the economic and epidemiological effects of different strategies for control of paratuberculosis in dairy herds. Animals occupy one of the six defined infection states; the spread of Johne's disease is modeled with five infection routes. Many different dairy farm situations can be simulated. Control strategies that can be simulated are: (1) test-and-cull; (2) calf hygiene management; (3) vaccination and (4) grouping of animals. Losses are caused by: (1) reduced milk production; (2) diagnosis and treatment costs; (3) lower slaughter value of cows and (4) sub-optimal culling. The benefits were calculated as reduction in the losses caused by Johne's disease; the costs of each strategy were calculated on the basis of actual costs of each item; and net present value (NPV) was calculated as benefits minus costs. Herd and prevalence data from The Netherlands and Pennsylvania, USA were used. In both situations, a low true mean prevalence within 20 years could be reached only when all calf management tools were applied. The Dutch control program (PPN) was on average economically attractive (with or without labor costs, the average NPV was Euro 1183 and 12,397, respectively). In Pennsylvania, contract heifer rearing and improved calf hygiene reduced the prevalence effectively and had large economic benefits (US$ 43,917 for 20-year period) if the calves were sent to the heifer facility while very young. Validation with data from 21 infected Dutch dairy farms (as well as face-validation: comparison of the results of the JohneSSim model with experiences of Johne's experts) supported the basic assumptions in the model.