Despite the press this has had and the evocative language that often appears in association with this question ('deadly' on the predatory size and inevitably 'skulking' on the scavenging one) the truth has been known to scientists for some time – they did both. Certainly tyrannosaurs scavenged when food was available, but they were clearly capable predators (or at least there was no truly strong evidence to suggest they were not) and some tentative evidence pointed towards animals that might have been injured by them.
Carnivorous dinosaurs actually shed their teeth quite often – like sharks they continually grew new teeth and the old ones would eventually fall out. So while finding shed teeth in or around carcasses of other dinosaurs is quite common, and in other cases we have found teeth embedded into bones, the question is how do we know this didn't happen when scavenging a carcass, but was a real strike on a living animal?
In the study by DePalma and colleagues, the tooth is buried deep within the mass of bone and is completely covered by it. The shape and texture of the bone growth is indicative of a major injury occurring to the hadrosaur and the obvious and entirely reasonable interpretation for this is that a bite from a tyrannosaur left the tooth in there. It's unlikely that this was an accident: large predators don't usually bite prey species very hard for no reason (and indeed probably rarely get close enough to do so except when going after them) so it's entirely justifiable to chalk this up to an attempted attack. "Attempted" is the key word here: the animal survived the incident and lived for many months or even years based on the amount of bone that had built up over the tooth.
This is also not a major surprise, few predators are successful much more than about half the time they try and hunt prey and of course many failed efforts would leave only a slight graze that would not show up on the bones, or be so bad as to kill the animal perhaps a few hours or days later where healing may not show. This doesn't show that Tyrannosaurus was a poor predator, merely that this hadrosaur got very lucky but lived to honk the tale for some time at least. Perhaps we are luckier still to find such an event marked in the fossil record, but this is hopefully the final, final nail in the 'scavenging only' idea and we can move onto some more detailed analysis of the behaviour of these giant carnivores.
R.A. De Palma II, et al., 2013. Physical evidence of predatory behavior in Tyrannosaurus rex. Proceedings of the National Academy of Sciences. doi:10.1073/pnas.1216534110
Note : The above story is reprinted from materials provided by guardian.co.uk. The original article was written by Dr Dave Hone