In recent years, concern about a new epidemic of retinopathy of prematurity (ROP) has focused attention on the increasing incidence of the disease and the factors responsible for its most severe consequences. Two studies designed to address these issues were done using data from three sources: the B.C. Health Surveillance Registry (Registry), Physicians's Notices of Livebirth (PNOB), and the Vancouver General Hospital (VGH). In the first study, Registry and PNOB records were used to determine crude annual birth weight-specific incidence rates for ROP in infants liveborn in the Province of British Columbia (B.C.) in the period 1952-1983. These rates showed that, in B.C., the original epidemic of the disease ended in 1954. Linear regression lines fitted for each of four birth weight categories showed that, in the 29 year period after 1954, there was a significant increase in the incidence of ROP-induced blindness in infants weighing less than 1000 grams at birth. To refine this observation, the data were sub-divided: the 29 year period, to two smaller periods, 1955-1964 and 1965-1983; the less than 1000 gram birth weight category to two sub-categories, 500-749 and 750-999 grams. Since the inter-period incidence should have been similar if the birth weight-specific incidence had not changed since the end of the original epidemic, the crude weight-specific rates for ROP-induced blindness in the early period were used to calculate the expected number of cases in the later period. When weight-standardized incidence ratios (SIR's) and 95% confidence limits were calculated, the results showed that, in the 750-999 gram sub-category, the SIR was significantly increased. Infants born in the period 1965-1983 were 3.07 times more likely to be ROP: blind than their equal weight counterparts in the earlier period. In infants weighing 500-749 and 1000 grams or more, there was no evidence to suggest an increase in incidence after 1954. The second study was done to determine the cofactors that differentiate infants who are blinded by ROP from those who are not. Infants were included if (i) they were born in B.C. between 1955 and 1983, (ii) they were known to the Registry as being ROP: blind (cases) or not blind (controls), and (iii) they were born in or admitted to the VGH within 28 days of birth. When the data from all three data sources were dichotomized and analyzed using univariate techniques, two variables, respiratory distress syndrome (RDS) and neonatal weight loss, showed a significantly protective effect. The effect of RDS disappeared when the data were stratified by birth interval indicating that the observed association was confounded by time. When the variables were reanalyzed in continuous form, none were significantly associated with visual outcome. However, since the power of the cofactor study was extremely low, none of the variables that were included can be eliminated as potential cofactors for the induction of blindness in infants with ROP. / Medicine, Faculty of / Population and Public Health (SPPH), School of / Graduate