Laser surface treatment, more specifically laser - (re)melting, -alloying and -dispersing, are techniques for improving wear, fatigue and erosion resistance of mechanical parts, using high power lasers. Analytical models which decrease these processes in a simplified way can be helpful for (a) reducing the time required to find optimum process parameters, (b) understanding major relations between process parameters and process results, and (c) the design of a real time process control system. In this paper a (quasi-stationary) analytical process model is presented, which relates the depth of the melt pool (in case of laser - remelting, -alloying and -dispersing), to the laser power and the relative velocity of the laser beam to the sample (feed rate). The model accounts for the latent heat of fusion and the energy produced in (or taken from) the melt pool, by exothermic (endothermic) reactions within the melt pool. The model shows a linear dependence of the melt pool depth on laser power and an inverse dependence on the square root of the relative beam velocity. The model was validated by experiments on Ti6Al4V alloyed with nitrogen, which is an exothermic reaction yielding TiN, and experiments on AISI304 alloyed with pre-placed chromium. Good correspondence between the experimental results and the model predictions indicates that the simplifying assumptions made in developing the model are justified within the range of application of the model.