While I agree that is the appropriate question to be asking, it's worth considering the length and complexity of the drug pipeline for creating new legal drugs,

While I agree that is the appropriate question to be asking, it's worth considering the length and complexity of the drug pipeline for creating new legal drugs, as a way of estimating the risk of new illegal drugs, which go through nothing like the same degree of testing: http://www.pacificbiolabs.com/drug_stages.asp http://www.pacificbiolabs.com/drug_stages.asp A new legal drug in its first iteration frequently through animal models frequently has a breakdown of 80% or more effectiveness, with the balance of the population mostly experiencing no effect and a small remaining fraction experiencing negative side-effects. The drug will then be tweaked and twisted to increase the size of the population it is effective on and drop the size of the negative side-effect population to as close to zero as possible. This is a hugely expensive process. It is where a good fraction of the hundreds of millions of dollars that go into the development of legal drugs wind up, and it is necessary because biochemistry differs very significantly across individuals. What the makers of illegal designer drugs appear to be doing is more like a developer who says "works on my machine" and then ships. No testing, no QA, no coverage... just ship based on a single working instance in a particular environment. That's a recipe for disaster, and while the disasters are mostly recoverable, the process is fraught with risk. This does not mean that making it illegal will change anything, because near as we can tell from the data the law barely changes human behaviour. Making abortion illegal does not change it's rate but just makes it unsafe. Making murder legal would not result in a vast increase in the rate of murders--it would result in blood-feuds and other mechanisms that the codification of law was designed to prevent. People don't refrain from murder because it's illegal (for the most part) but because they believe it's morally wrong. The law follows from that widespread moral belief, not the other way around. So legalization and regulation may well be the best way to go, but given the ordinary standards of drug development that might create an underground grey-market where things are almost as bad, safety-wise, as they are now. A carefully targeted criminalization, that focuses solely on drug makers rather than distributors or users, might be a better solution. So long as we keep in mind that the goal is to minimize harm, including the harm done by law enforcement.