Coffee brew presents sensory, stimulatory and antioxidant properties highly appreciated by consumers, despite being associated with an increase in the level of blood cholesterol due to the effects of the diterpenes, especially cafestol, present in the lipid fraction. Although it is believed that the paper filter retains the brew diterpenes, new studies have shown that sometimes coffee filtered through paper can also increase the blood cholesterol level, putting in doubt the efficiency of the paper filter in retaining the diterpenes. Thus the objective of the present study was to verify the distribution of cafestol between the paper filter, the spent coffee and the coffee brew itself, from two coffee samples containing high and low cafestol contents selected from 13 samples of different cultivars and from different locations. In addition, the effect of the roasting degree on the cafestol contents of the roasted coffee was evaluated and the relationship between particle size of the roasted coffee and the extraction of solids. The highest cafestol content was found in the lightly roasted coffee, and the coffee brew presented higher solids contents when the particle size of the coffee powder was below 500 μm. The results showed that of the initial cafestol concentration present in the roasted coffee, the paper filter retained 12.41%, the spent coffee 87.45% and the brew 0.15%. Thus, one can conclude that the greater part of the coffee cafestol is retained by the spent coffee, due to the low extraction of the lipid fraction by the hot water.