Blocking the direct membrane immunofluorescence reaction, obtained by exposure of cells carrying Epstein-Barr virus to the fluorescein-isothiocyanate-conjugated reference IgG of a Burkitt's lymphoma patient (Mutua), wds evaluated in relation to variou's test conditions. Membrane antigen expression, assessed by the frequency of positive cells, and conjugate dilution did not significantly affect, within the set limits, the blocking activity of undiluted and serially diluted sera, except for a small decrease in the activity of low-blocking sera with increasingly reactive cells. The tests showed good reproducibility, but 1 of 5 observers recorded higher blocking activities than the other observers in parallel tests. From repeated blind tests, confidence intervals were obtained for the mean blocking activity of one serum dilution. Twofold dilutions of reactive sera give a linear decrease of blocking activity within certain ranges. This can be used for the titration of membrane-reactive antibodies. An arbitrarily chosen blocking level (0.5 or 0.4 blocking index units) may serve to define the titer endpoint. Evaluation by the same observer of repeated titrations over 5 two-fold dilution steps is recommended, with all sera from the patient to be tested included each time. Adaptation of the observer to the reaction patterns in each test by reading reference cell samples after no, low, and high blocking of the reactions gives higher accuracy.