An improved methodology has been developed to characterise running-in and steady-state wear processes. The experimental study was conducted with En 31 steel specimens on reciprocating tester with ball-on-flat geometry under lubricated sliding conditions. The tests were conducted according to a factorial design. The variables selected were the load, temperature and surface roughness. The wear behaviour for a given set of operating conditions has been characterised on the basis of developed methodology. The parametric influence of operating conditions was then analysed on the basis of polynomial relationships. The steady-state wear was found to be significantly influenced by the initial roughness.