The relationships between 22 variables and life satisfaction are explored, testing in particular for partial effects and determining the extent to which each predictor makes an independent contribution to explaining perceived satisfaction. Based on data from 274 females and 233 males aged 45 and over, a number of important relationships are noted. Holding socioeconomic status constant, generally observed relationships such as between life satisfaction and age, marital status, and family size disappear. Other relationships found in previous research are reduced in magnitude by the introduction of the control variables. Computing standardized partial beta coefficients, it is found that the best predictors of life satisfaction are socioeconomic status, perceived health status, and informal participation with nonlcinsmen. Neither interaction with kin nor most of the social and personal background characteristics considered contribute significantly to explaining reported satisfaction. The implications of these findings for future studies and theory construction efforts are suggested.