Baseline data from a study run in 1974, which comprised beta 2-microglobulinuria (beta 2mu) measurements at monthly intervals from 416 members in 112 households from the BEN affected village of Petka, were compared with the results of two subsequent, cross sectional studies. In June 1988, retesting involved 320 available persons from the same households. Another collection of 284 urine specimens took place in October 1989. Prevalence of tubular proteinuria was the same in 1988 and 1989 as it was in 1974, indicating that the level of exposure to nephrotoxic agent did not change over time. Over 94% of the individuals who were always beta 2mu negative in 1974 remained negative in 1988. By contrast, over two thirds (68.7%) of those who were positive two or more times 14 years ago, tested positive upon re-examination in 1988. Particular interest arises from the data on those initially repeatedly positive persons in whom the overt disease did not occur over time; moreover, some appeared to be unaffected in 1988 and 1989 according to our set of laboratory criteria. The results suggest occasional slow progression and even possible reversibility of tubular lesions in individuals living in the BEN affected environment.