I believe @madman has linked to studies that are far less charitable. My own results are pretty striking in comparing LabCorp's direct immunoassay test to the Tru-T calculation. Free testosterone is expected to be pretty linear with dose. It's true that N is small in these data and the exceptionally good correlation of Tru-T is surely overstated. Nonetheless, it's food for thought....
I know dialysis is the method always championed on this website, but science seems to believe the different test methods correlate very highly (r > 96).
...