Ill agree with Julian , ELISA is highly sensitive , good specificity and does not involves radioisotopes. Also good secondary antibody selectivity. Overall the technique is accurate. RIA also had advantages but as Julian mentioned theres lot of safety issues that practically made them not so common to be used these days. One other disadvantage was the short shelf- life of the radioactive compounds.
My impression is that RIA will be more accurate by EIA will be more convenient, since no handling and disposal of radioactive materials will be required.
Nobody is doing RIAs anymore due to safety issues and regulations. As the detection limit of ELISAs is not inferior, I recommend to stick to this method.
Julian Baumeister, "nobody" is definitely overstatement. Remember, quantifier of universality makes every statement false :).
ELISA involves enzymatic step which definitely adds an experimental error. Hence, RIA must be more accurate. Regulations are indeed an issue (real safety is not, of course, if an equipped isotope lab is available), but regulations are different in different countries. So, generally, if you have a radioisotope lab and it is possible in your country to do RIA without too much bureaucracy, I would recommend RIA.
Ill agree with Julian , ELISA is highly sensitive , good specificity and does not involves radioisotopes. Also good secondary antibody selectivity. Overall the technique is accurate. RIA also had advantages but as Julian mentioned theres lot of safety issues that practically made them not so common to be used these days. One other disadvantage was the short shelf- life of the radioactive compounds.