The purpose of this study was to examine mortality trends for valvular heart disease in Canada. Mortality data and population data were obtained from Statistics Canada publications and analyzed according to the diagnostic criteria established by the International Classification of Diseases (ICD) of the World Health Organization. Because of the changes in ICD classification over the years, attempts were made to analyze data only when a relevant and comparable diagnostic class was available. Mortality data were collected over time, therefore they tended to be strongly autocorrelated and statistical methods were utilized to minimize this factor. In addition, statistical methods for switching regression models were applied to evaluate trends for possible changes. For chronic rheumatic heart disease ICD 393-398, there was a significant trend of a decrease in death rate of about 0.22 per 100,000 population per year. These trends were observed for both men and women although the death rates for women were consistently greater than those for men. Switching regression models suggested a switchover point, but only for men, indicating a flattening or decrease in the downward trend after 1978. For rheumatic mitral valve disease there was a significant decrease in death rate of 0.026 deaths per 100,000 per year and for rheumatic aortic valve disease of 0.036 per 100,000 per year. For nonrheumatic mitral valve disease there was a significant decrease in death rates of 0.01 per 100,000 per year, but there was no trend for nonrheumatic aortic valve disease.(ABSTRACT TRUNCATED AT 250 WORDS)