We approximate the distribution of scan statistics generated by some Bernoulli and Gaussian stationary 1-dependent models. We compare the results with the corresponding results provided by i.i.d. and Markov models.