Take "e" to the power of the coefficient, where "e" is the base of the natural logarithms, approximately 2.718. Many calculators have an "e^x" button. This is the odds ratio. For example, if the coefficient is 3.21, then the odds ratio is e^3.21 = 24.78. It is easier to interpret the odds ratio than the coefficient.
Determine if the coefficient applies to a dichotomous, categorical or continuous variable. Each coefficient is related to one independent variable. That variable can be dichotomous (e.g. sex), categorical (e.g. race/ethnicity) or continuous (e.g age).
Interpret the odds ratio for dichotomous independent variables. This is the odds of a particular outcome for one level of the independent variable divided by the odds of that outcome for the other level. For example, if the odds ratio for females voting for Obama is 2.3, that means that the odds of a woman voting for Obama are 2.3 times those for men.
Interpret the odds ratio for categorical variables. With categorical variables, one level will be the reference level and will have an odds ratio of 1 (or this may be left blank in the computer output). The odds for other levels of the categorical variable are relative to this. For example, if ethnic group is classified as "Caucasian," "African-American," 'Asian," "Native American" or "Other," and "Caucasian" is the reference group while "African-American" gets an odds ratio for voting for Obama of 3.2, it means that the odds of an African-American person voting for Obama are 3.2 times those of Caucasians.
Interpret the odds ratio for continuous variables. Here the odds ratio is per unit of the independent variable. For example, if the odds ratio of age for voting for Obama is .98, it means that the odds of a 23-year-old voting for Obama are .98 times those of a 22-year-old, and similarly for every year of age.