Purpose: The purpose of this paper is to measure the relative effect of relevant explanatory variable on smoking tendency and smoking intensity. Design/methodology/approach: Using survey data collected by the Israeli Bureau of Statistics in 2003-2004, a probit procedure is estimated for analyzing factors that affect the probability of being a smoker. Using ordinary least square methods, the factors that affect smoking intensity are also estimated. Findings: The findings show that the probability of being a smoker, as well as having greater intensity of smoking, is larger for males, non-Jews, and those who are un-married and younger. When estimating different equations for males and females, the results for males are in the same direction as the results for all sample observations, but for females it is found that being Jewish significantly increases smoking probability and females born in America or Europe smoke more intensively. Research limitations/implications: Our findings enable the focusing of the relevant authorities' attention on the population that is at a higher risk of smoking. Practical implications: Concurrent with the necessary legislation, a strategy has to be developed that can appropriately target the different educational, ethnic, age and gender groups, and effectively communicate the correct health message that will properly impact on long-term behavior. Originality/value: The findings contribute to existing knowledge in two main aspects: the first is in measuring the relative effect of each significant explanatory variable on smoking tendency and smoking intensity. The second is in using a unique data set that includes many agents' characteristics and examining the significance of the various characteristics on smoking tendency and smoking intensity. (Contains 15 tables, 1 figure, and 6 notes.)