The question considered in this review is the extent to which changes in the prevalence or severity of enamel fluorosis have occurred over the last half-century. Emphasis is given to a review of those studies in which subjects are drinking water that is fluoride deficient and those in which subjects are drinking optimally fluoridated water, either adjusted or natural. Trends in fluorosis were examined using two definitions of fluorosis (definite and any signs) and three types of comparisons--comparisons of pooled estimates from all available studies that include data from different communities and time periods, comparisons of estimates from the same communities at different times, and comparisons of estimates from selected studies in the early years of fluorosis research with results of the US National Fluorosis Survey done by the National Institute of Dental Research. A clear increase in fluorosis among populations drinking community water that contains less than 0.3 ppm fluoride was found. Results of the comparisons using studies with Dean's Index pooled at different time points, comparisons in the same communities over time, and comparisons of prevalence found in selected communities before fluoride was widely available with the National Fluorosis Survey all support this conclusion. An increase in the prevalence of fluorosis in those drinking optimally fluoridated water likely has occurred as well; however, evidence for such a trend is not as clear as for fluoride deficient communities because of mixed results depending on the type of comparison. The majority of fluorosis cases continue to be mild and seem of little esthetic consequence for most of the public or dental profession. But a few cases of more severe fluorosis can be found now in some communities. Because the prevalence of fluorosis is now higher than 50 years ago, we can conclude that fluoride availability to the developing enamel during critical periods when enamel is at risk of fluorosis has increased in North American children.