In 1990 and 1992, in previous reviews of the literature and in reports of their experience with both amyloid and non-amyloid monoclonal immunoglobulin deposition diseases, the authors proposed a classification scheme encompassing all the forms of non-antibody-mediated monoclonal immunoglobulin deposition. The premise underlying the proposal was that the mode of pathogenesis of each of the various disorders is similar. Monoclonal expansion of a B-cell and plasma-cell population producing an excess immunoglobulin polypeptide with structural characteristics predisposing to tissue deposition in either the fibrillar or nonfibrillar state would be associated with organ-compromising deposits in tissue. At that time it appeared that LCDD and LHCDD were more likely to occur in the course of myeloma in which the other features of the neoplastic cells (i.e., marrow suppression, lytic lesions, recurrent infections) were also clearly evident. At this time, the authors' additional experience suggests that this judgment may have been premature, based in part on too small an initial sample and in part on the use of diagnostic criteria for multiple myeloma that may have not been sufficiently precise. The authors now believe that the nodular glomerulopathic form of NAMIDD is similar in both course and prognosis to AL amyloidosis occurring in the absence of multiple myeloma (primary amyloidosis). The primarily tubular basement-membrane form of the disease usually seen with concurrent myeloma kidney with BJCN, is associated with more aggressively proliferative plasma-cell neoplasms. The authors believe that these associations relate to the size of the malignant clone which, in turn, determines the amount of depositionogenic protein available and the rate of its presentation to the target organ (primarily the kidney). The distinction is not trivial, for if the authors are correct, their data suggest that not all forms of renal disease occurring in the course of plasma-cell dyscrasias have the same bleak prognosis. The outlook for nodular glomerular disease, as an indirect marker of clone size, may be intrinsically better than that of a renal biopsy showing cast nephropathy and tubular basement membrane LCDD deposits and clinical renal failure. Since 1992, it has also become less certain that there are general structural differences between light chains forming amyloid and those producing non-Congophilic tissue deposits. The current data suggest that light-chain proteins with the capacity to form pathogenic tissue deposits may exist in a spectrum, with one end represented by those only capable of forming amylord, the other by those depositing in a more amorphous, nonfibrillar manner, and a group in the center capable of either or both, depending on circumstances that are presently not understood. An alternative view suggests that all or most proteins depositing as fibrils pass through a non-Congophilic, nonfibrillar phase, of a length varying according to their primary structure, which is not detected in vivo because of the vagaries imposed by clinical sampling. More structural analyses of material extracted from deposits in tissue may resolve this issue.