The excursion set approach has been used to make predictions for a number of interesting quantities in studies of non-linear hierarchical clustering. These include the halo mass function, halo merger rates, halo formation times and masses, halo clustering, analogous quantities for voids and the distribution of dark matter counts in randomly placed cells. The approach assumes that all these quantities can be mapped to problems involving the first-crossing distribution of a suitably chosen barrier by random walks. Most analytic expressions for these distributions ignore the fact that, although different k-modes in the initial Gaussian field are uncorrelated, this is not true in real space: the values of the density field at a given spatial position, when smoothed on different real-space scales, are correlated in a non-trivial way. As a result, the problem is to estimate first crossing distribution by random walks having correlated rather than uncorrelated steps. In 1990, Peacock & Heavens presented a simple approximation for the first crossing distribution of a single barrier of constant height by walks with correlated steps. We show that their approximation can be thought of as a correction to the distribution associated with what we call smooth completely correlated walks. We then use this insight to extend their approach to treat moving barriers, as well as walks that are constrained to pass through a certain point before crossing the barrier. For the latter, we show that a simple rescaling, inspired by bivariate Gaussian statistics, of the unconditional first crossing distribution, accurately describes the conditional distribution, independent of the choice of analytical prescription for the former. In all cases, comparison with Monte Carlo solutions of the problem shows reasonably good agreement. This represents the first explicit demonstration of the accuracy of an analytic treatment of all these aspects of the correlated steps problem. While our main focus is on first crossing distributions of deterministic barriers by random walks, in Appendices we also discuss several issues that arise upon introducing a stochasticity in the barrier height, a topic which has gained interest recently with regard to the mapping between first crossing distributions and halo mass functions.