As with all binaries, those that contain massive stars reveal various degrees of interaction, depending mainly on orbital separation and age, although things happen much faster in massive binaries. Those massive binaries with initial periods exceeding ~10 years generally only interact via wind-wind collisions, with little or no effect on their subsequent evolution (unless located in dense clusters). Shorter-period systems show even stronger wind-wind collisions as a rule, but also interact more directly via Roche Lobe Overflow or Common Envelope, with dramatic effects on their evolution. If we didn't have binaries among massive stars, we would be missing a whole host of interesting phenomena in the Universe, such as sources of enhanced stellar X-ray or non-thermal radio emission, WR dust-spirals, inverse mass-ratios, very rapid spin, rejuvenation and massive blue-stragglers, enhanced cluster dynamics, many runaways and possibly even SMBHs and GRBs! On the other hand, non(or little)-interacting massive binaries are also useful to provide information on Star-Formation processes and determination of stellar parameters (such as the mass) that would otherwise be difficult or impossible to obtain from single stars. In this review, I highlight some of the developments that have occurred during the past few years since the last IAU Symposium on Massive Stars in 2002.