We measure radiation dose in units called rem. Above 1,000 rem, incapacitation occurs within minutes, followed by extreme fatigue and nausea, then death. For smaller doses, 300 to 500 rem, about half of the victims die within a month.
But reduce the dose a little more, down to 100 rem, and the effects are very mild. At 50 rem, nobody even gets sick.
This 'threshold effect' creates problems for the terrorist. He can put radioactive material into a bomb - but concentrated radioactivity kills fast, so he'll have to protect himself with a ton of lead.
Then he must deliver the bomb, take it out of the lead shield and explode it with dynamite so the radioactive material spreads into the air.
He wants the wind to carry it around the city, but has to hope it doesn't spread too much or it won't kill anyone.
If radioactivity comparable to that of the Goiania accident were spread over a square mile, the exposure to people - even if they stayed outdoors for a month - would have been only 10 rem, far below the lethal threshold. There would be no radiation illness, no dead bodies.
Low levels of radiation do cause harm, but it is hidden and delayed: tiny mutations in your DNA that increase your risk of cancer by a very small amount.
Would a terrorist be satisfied with the knowledge that his bomb might cause a small increase in cancer deaths among a small number of people in many years' time? ... To my mind, the real terror weapon of choice is the one they can buy at the local petrol station.