Low density lipoprotein (LDL) particle size is inversely associated with plasma triglyceride concentration in cross-sectional analyses. In the present study, changes in the LDL particle size of 227 participants of the Framingham Offspring Study were analyzed longitudinally by nondenaturing gradient gel electrophoresis at two examinations that were separated by 3-4 years. All subjects had triglyceride concentration <400 mg/dl at both exams. Using laser scanning densitometry to assess mean LDL particle size, 56% of samples displayed a change in size: 41% had a one-band size change, 13% had a two-band change, an 2% had a three-band change. These changes in size corresponded to a 15% change in pattern type, based on pattern A and B terminology. There was a significant inverse association between change in LDL size and change in triglyceride (p<0.0001) and glucose (p<0.004) concentrations, body weight (p<0.02), and age (p<0.03). There was also a significant positive association with change in high density lipoprotein (HDL) cholesterol concentration (p<0.0001). Change in LDL cholesterol concentration, as calculated by use of the Friedewald formula, however, showed no significant association with change in LDL size (p<0.9). There was also no significant association with change in smoking or blood pressure, but there was a nonsignificant inverse trend associated with alcohol intake (p<0.08). Women had more significant changes than men, but the relation between change in LDL size and change in lipids was the same for women and men. In stepwise regression analyses of biochemical indices, change in triglyceride concentration produced the best single-variable model (r=0.460). When change in HDL cholesterol concentration was added to the model, the r value increased slightly to 0.481 for the two-variable model. No other variables entered the model. We conclude that the amount of circulating triglyceride appears to be the single most important factor affecting LDL particle size and that LDL size can be modulated by changes in plasma triglyceride concentration.