Treatment of porcine lutropin beta-subunit by increasing amounts of dansyl chloride shows that only one fluorescent group can be bound to the free subunit, namely on tyrosine beta-37. This modification prevents reassociation with native chi-subunit. In contrast to dansyl chloride, dinitrofluorobenzene reacts preferentially with tyrosine beta-59. This substitution does not interfere with the reassociation with the native chi-subunit. By using equimolar ratio of dansyl chloride and porcine lutropin chi-subunit it is possible to modify this subunit on a single site, which is found to be a tyrosine residue. The monodansylated X-subunit is still able to recombine with native B-subunit but its fluorescence is found to be markedly quenched upon binding. In addition, the O-dansyl-tyrosyl fluorescence quenching by potassium iodide is more effective on the recombined dimer than on the free chi-subunit. Both reconstituted dimers (native xhi chi dinitrophenylated beta and dansylated chi chi native beta) are without biological activity. It is not clear whether the substituted phenolic groups are essential or whether the added groups prevent the chi beta dimer from taking the active conformation. This alternative is discussed in the light of recent data from this laboratory and others.