Plasma fibrinogen was determined in 189 twins participating at the Indiana center during the third examination of the NHLBI twin study with a mean age of 63 years. Moderate heritability estimates were obtained from 44 complete MZ pairs and 39 complete DZ pairs. After adjustment of fibrinogen levels for age and other confounding variables related to cardiovascular disease risk, the maximum likelihood heritability estimate was only 30% (p = 0.03). Plasma fibrinogen was most strongly associated with smoking and the presence of diabetes. Omitting all subjects with diabetes or cardiovascular disease further reduced the heritability estimates slightly, and most path models including genetic parameters provided no significant improvement in fit over a model determined solely by random environmental effects. Our results are consistent with the environment rather than genetic influences having a greater influence on the level of plasma fibrinogen.