Chromosome monitoring of 500 batches of WI-38 cells prepared for vaccine manufacture involved examining 50-60,000 metaphases at p31-32. Results collected over 6 years showed the mean incidence of aneuploidy and breaks to decrease slightly, whereas the incidence of structural abnormalities followed a sharply rising trend, eventually reaching a level where a substantial proportion of cell batches exceeded the limit set for this parameter for vaccine substrates. No technical or cultural factor accounted for the trend which appeared to be correlated with the date of receipt of the seed ampoule. The confidence limits set for this parameter in 1971 were regarded then as provisional. It is suggested that in the light of the observations reported, the position should now be reconsidered.