I'd say length-of-action is just as big a factor, especially with prescription pain-killers. When people take drugs with short half life's, they tend to 'dose up'. Which, if you're not careful can easily lead to addiction.

Also, I do believe personality comes into play. When I was taking the max dose of Tramadol for over a year, I stopped dead three weeks post-op. The only noticeable side-effect was a runny-nose for a few days, hardly a big deal.

I just mean in terms of potential addictiveness. Mind you, weakly binding drugs will usually be coupled with higher doses. So there's not exactly a separation between time-to-action and length-of-action and binding affinity. Dont worry about coatings and whatnot. If someone wants to abuse a drug, it will most likely be crushed up. Now, if we are talking about the actual substance - depending on whether it is a direct agonist versus a prodrug that goes through multiple passes, or has other rate limiting behavior. -- This will definitely affect the addictiveness of the drug. But 100% in addition to the binding affinity and the receptors affected. So I agree, it's a combination of a bunch of factors. But the ultimate end result, which is perceived duration, perceived impulse, is responsible for a large part of the addictive threshold of that drug. The drug's makeup is clearly responsible for everything, I'm not saying perception trumps the actual chemistry. I just mean, if there is very little perceived impulse, there will most likely be little to no physical dependency. Again, this is based on the tendency of someone to use a drug at what we want to call effective doses, or in abusing it, supra-effective doses.