To measure the effect of arthritis and musculoskeletal conditions on working life expectancy.Cross sectional data from the 1994 Canadian National Population Health Survey (NPHS) were used to calculate and compare the working life expectancy of individuals who reported "arthritis or rheumatism" with that of the general population. Age and sex-specific workforce participation rates were calculated for the population reporting arthritis or rheumatism as a chronic condition, excluding back pain, and for the entire population surveyed. Age and sex-specific population figures and mortality data were obtained from annual estimates produced by Statistics Canada. Working life expectancy was estimated by constructing multiple-decrement life tables for the total and for the arthritis and rheumatism populations.The NPHS surveyed 22,000 households, yielding a sample size of 58,439 individuals. The percentage of the population aged 15 to 65 yrs who reported having arthritis or rheumatism was 8.9%. The percentage of persons employed for each group was reduced compared to the total population, by 3 to 23%. Working life expectancy of individuals with arthritis or rheumatism was reduced by 4.19 +/- 0.02 yrs (mean +/- SE) for men and 3.12 +/- 0.01 yrs for women at age 15 (p < 0.001 for both), with a persistent reduction through all age groups. Working life expectancy of men at age 15 was 37.42 +/- 0.01 yrs for the population with arthritis or rheumatism compared to 41.62 +/- 0.01 yrs for the total population; for women it was 31.06 +/- 0.01 and 34.19 +/- 0.001 yrs for both groups, respectively.The working life expectancy of people with arthritis and musculoskeletal conditions is significantly reduced compared to the general Canadian population.