We use the Monte-Carlo (MC) test technique to find valid p-values when testing for discontinuities in jump-diffusion models. While the distribution of the LR statistic for this test is typically non-standard, we show that the MC p-value is finite sample exact if no other (identified) nuisance...