The capture and gradual inspiral of stellar mass objects by a massive black hole at the centre of a galaxy has been proposed as one of the most promising source of gravitational radiation to be detected by LISA. Unfortunately rate estimates for this process suffer from many uncertainties. Here we report on the use of our newly developed Monte Carlo stellar dynamics code to tackle this problem. We present results from simple galactic nuclei models that demonstrate the high potential of our approach and point out the aspects of the problem where an improved treatment seems desirable.