During the last years, a trend to replace commonly used short arc lamps in projection systems with alternative light sources is seen. Next to LEDs for low light output products, lasers try to enter the projection area and have the ambition to infiltrate from low (picoprojection) towards high light output systems (digital cinema).One of the benefits of lasers is their narrow spectral bandwidth. As a consequence, the display can have a very large colour gamut, if the lasers are carefully selected. Another benefit is the very low intrinsic étendue of the source. One can imagine using less complex, more efficient, smaller but more powerful optical systems. This not only for scanning projectors, but also for 2D light valve based projectors (LCOS, LCD, DLP). In addition, the limited lifetime of lamps has serious impact on a system's cost of ownership, and puts light source reliability/lifetime high on the list of priorities for future developments. For this reason, Barco entered the European FP6-project OSIRIS, in a subtask where a 300 lm laser projector demonstrator has to be developed and evaluated.So far, we found out that next to obvious challenges such as laser cost and laser power, the most critical issue regarding image quality is speckle interference which counteracts the beneficial nature of the light source. This phenomenon is a direct consequence of the coherent nature of a laser and cannot be solved as easily as is often claimed. We will describe laser speckle in the context of laser projection and the theoretical limits of several reduction techniques. This leads to guidelines which can make laser projection worth considering.