Before they can sell drugs, pharma firms need to make them. And before making them, they need to figure out what to make. That’s both time-consuming and expensive – so if you’re going to invest in the drugs business, you need to balance the time spent on all the failures with the potential of the rare success.

Drug discovery involves toiling away in a lab to first understand what causes a disease, then to figure out which molecules can stop it. Pharmaceutical companies do some of this work themselves, but they also fund university researchers to do the rest of it for them. That’s because of the sheer amount of research there is to be done: 5,000 to 10,000 different molecules will need to be tested for any one drug. The initial process alone typically takes around a decade, and costs around $500 million.

Eventually, the pharma firm will have shortlisted a few molecules it reckons might work, and that’s when the long process of clinical trials begins. Because before a drug can be approved for use – by the Food and Drug Administration in the US, and the European Medicines Agency in the EU – it has to pass three phases of clinical trials to check it’s safe and effective.

The first phase takes around a year, and involves confirming the drug’s safety and dosage with a very small pool of people. The second takes two or three years, and involves testing the treatment’s effectiveness on a few hundred human guinea pigs. If a drug makes it through both of those stages without revealing any nasty side effects, the third and final phase assesses thousands of test subjects to help evaluate the drug’s long-term impact.

Most drugs – 85-95%, to be specific – don’t make it through clinical trials. That means there’s a lot of trial and error involved in drug development, which is why it’s so costly. Estimates vary for exactly how much drug development costs from start to finish, but the (pharma-funded) Tufts Center for the Study of Drug Development says it’s an average of $2.6 billion. And $1.2 billion of that is what’s known as “opportunity costs” – the amount investors missed out on by not investing their cash elsewhere.

Others think $2.6 billion is too high. Harvard’s School of Public Health, for example, says Tufts uses “assumptions highly favorable to the industry” so they can justify their high drug prices, and that “the actual number is probably much smaller”. Either way, we’re still talking a lot of money to produce one drug. And because lots of the most easily cured illnesses have already been solved, drug development is only going to get harder and more expensive.

Still, you’d think that once they have developed the next blockbuster drug, the pain would be over for pharma firms. Manufacturing costs might be increasing (more on that later), but they’re still negligible compared to the costs of development. But no: turns out marketing a drug’s pretty pricey too…

The takeaway: Drugs can cost billions to develop, and the process can take decades.