Objective: The determinants of adult obesity were examined using data from a large-scale health survey that included information on socioeconomic status, education level, employment status, healthy behavior, subjective health perception, and emotional factors. Through adequate weight control education, it is expected that the prevalence rate of obesity can be lowered. Methods: A multiple logistic regression model was used to analyze data extracted from the 2001 Taiwanese National Health Interview Survey (NHIS 2001). Subjects were classified into two groups, obese and non-obese, by analysis of binary variables. Results: An inverse U-shaped curve relating to adult obesity was found for different age groups. Married males (OR=1.31, 95%CI: 1.05-1.62), females with a higher education level (OR=0.19, 95%CI: 0.04-0.42), people who consumed betel nuts (OR=0.73, 95%CI: 0.65-0.82), and males who were employed (OR=1.15, 95%CI: 1.09-1.24) revealed a higher likelihood of obesity, while people with a better subjective health perception (OR=0.84, 95%CI: 0.74-0.93), people with a low mood score (OR=0.98, 95%CI: 0.97-0.98), males who smoked (OR=0.65, 95%CI: 0.42-0.78), and females who were employed (OR=0.78, 95%CI: 0.64-0.99) demonstrated a lower likelihood of obesity. In terms of diet, people who habitually ate breakfast and people whose diet contained a higher percentage of meat revealed a higher likelihood of obesity, while females who ate more vegetables and those who exercised regularly demonstrated a lower likelihood of obesity. Conclusion: The results of this study endorse the need for the government to reinforce obesity education programs, especially in regards to diet and lifestyle modifications, particularly for those individuals with a lower level of education.