Even though the anticarcinogenic effect of dietary factors especially beta - carotene has been reported by various investigators, the mechanism of the action of ${eta}-carotene$ has not yet been identified. We carried out the present study to determine the possibilities of relative cancer risk related to dietary intake of vitamin A ( both ${eta}-carotene$ and retinol ) and blood levels of vitamin A among Koreans. The subjects were divided into two groups; cancer patients and controls. Blood levels for ${eta}-carotene$ and retinol were analyzed by alumina column chromatography and colorimetry. Dietary intake was examined by food profile and convenient method for evaluating nutritional status through recalling 10 years of food habits. The results obtained are as follows : 1 ) Calorie, protein, fat, and carbohydrate intakes of cancer patient were lower than those of control. Calorie and carbohydrate intakes showed no significant difference but protein and fat intakes were significantly lower in cancer patients. According to cancer sites, in stomach cancer only fat intake was significantly lower than that of control. In lung and larynx cancer calorie, protein, fat and carbohydrate intakes showed similar trend as in control. 2 ) Vitamin A intake of cancer patient was significantly lower than that of control. It was estimated that 83.6% of total Vitamin A intake were provided by ${eta}-carotene$ for control and cancer patient respectively. 3 ) The mean intake of dietary ${eta}-carotene$ in cancer patient was significantly lower than that in control ( $7002;mu}g/day$ versus $10326;{mu}g/day$ ) According to cancer sites in mean intake of dietary ${eta}-carotene$ was significantly lower in all but stomach cancer compared with that of control. Lung and larynx cancer showed lowest ${eta}-carotene$ intake with mean value of $5855{mu}g/day$ and $5492{mu}g/day$ respectively. 4 ) The mean intake of dietary retinol in cancer patient was significantly lower than that in control ( $245{mu}g/day$ versus $338{mu}g/day$ ), but the difference was not significant. 5 ) The relative risk of all cancers in the first (lowest) to the forth quartile level of ${eta}-carotene$ consumption such as $0-5999{mu}g/day$. $6000-8999{mu}g/day$, $9000-11999{mu}g/day$/ day and $12000-20000{mu}g/day$ was 85 : 1.7 : 20 : 1.0. The relative risk of all cancers in the first (lowest) to the forth quartile level of retinol consumption, such as $0-299{mu}g/day$, $300-599{mu}g/day$, 600-899${mu}g/day$, and $900-1200{mu}g/day$, was 1.14 : 067 : 0.21 : 1.0. 6 ) The various food group consumption of cancer patient were significantly lower than those of control in green leafy vegetables, fruits, sea weeds, milk and cheese and eggs. But the Kimchie consumption in cancer patient was three fold higher than that of control ( $1840;{mu}g/day$ versus $562;{mu}g/day$ ) and in the stomach cancer, Kimchie consumption was the highest, ( $1890;{mu}g/day$) According to cancer sites, the consumption of green leafy vegetables was significantly lower in all but stomach cancer compared to control and other vegetables showed no difference between two. In lung and larynx cancer, green leafy vegetables consumption was lowest ( $6094{mu}g/day$ $5921{mu}g/day$) and milk and cheese consumption was also( $5;{mu}g/day$ and $11{mu}g/day$) 7 ) The recovery of ${eta}-carotene$ from human serum by alumina column chromatography was $94.4{pm}2.3%$. 8 ) Cancer patients showed significantly lower serum retinol ($56.4{pm}18.1:{mu}g/100ml$ versus $72.2{pm}21.8:{mu}g/100ml$) and ${eta}-carot..