Time-series regressions corrected for serial autocorrelation were used to explore whether time-series correlates of the French suicide rate were dependent upon the use of crude versus age-adjusted rates for the period 1950-85. For males, divorce, marriage & birth rates all contributed significantly to the prediction of the crude & age-adjusted suicide rate. For females, only divorce rates contributed significantly. The signs of the 3 predictors were identical for predicting both types of rates.