Background: Energy-dense food and low physical activity have been blamed for the dramatic rise in the prevalence of obesity and related metabolic disorders. However, few dietary and physical activity factors have been consistently associated with obesity in observational studies. Self-reported behaviors are prone to misreporting, which may partly explain the inconsistency of previous results. Nutritional biomarkers provide an objective approach to measure habitual intake, but valid biomarkers are not available for all types of dietary exposures. Aims: The aims of this dissertation were to 1) validate self-reported dietary intake and physical activity by using objective co-twin comparison assessments (I) and the doubly labeled water (DLW) technique (II); 2) determine whether eating and physical activity behaviors are associated with body mass index (BMI) and waist circumference (WC) (I, II); 3) examine whether acquired obesity (IV) and dietary factors (III, V) are associated with serum lipoprotein profiles. Materials and methods: The following participants were recruited from the FinnTwin12 and FinnTwin16 studies, two population-based studies on young adult Finnish twins: 1) 713 monozygotic (MZ) and 698 dizygotic (DZ) twin pairs of the same sex who filled in food-frequency questionnaires (FFQ) and co-twin comparison questions (I); 2) 14 MZ twin pairs discordant (BMI difference more than 3kg/m2) and 10 pairs concordant for obesity who provided 3-day food and activity diaries, eating behavior questionnaires, co-twin comparison questions and total energy expenditure (TEE) measurements determined by DLW (II); 3) 15 obesity-discordant and 9 concordant MZ twin pairs who completed 3-day food diaries, physical activity questionnaires and measurements of subcutaneous adipose tissue, visceral adipose tissue and liver fat by magnetic resonance imaging and spectroscopy and of serum lipid determinations by ultracentrifugation, gradient gel electrophoresis and enzymatic techniques (III, IV); 4) 663 twin individuals who provided FFQs and nuclear magnetic resonance spectroscopy-derived measurements of serum lipoproteins and serum docosahexaenoic acid (DHA) as an objective biomarker of DHA intake (V). Results: In both zygosity groups, the co-twins for whom both twin pair members concordantly answered that the identified twin eats more (overall), or eats more fatty foods, or eats more sweet and fatty delicacies, or eats faster, or selects food less according to healthiness or makes less active choices in daily life had significantly higher BMIs and WCs than their twin siblings. Eating more (overall), eating more sweet and fatty delicacies and exercising less remained significant predictors of intrapair differences in BMI and WC independent of each other as evaluated by multivariate regression analysis. Co-twin comparison questions on snacking, fatty foods, sweet and fatty delicacies and healthy foods corresponded well with self-reported food intake in the FFQs. Twin pairs who differed in the overall amount of food they consumed had the largest intrapair differences in BMI (MZ: 1.9 ± 0.1, DZ: 2.9 ± 0.2 kg/m2) and WC (MZ: 5.5 ± 0.6, DZ: 7.5 ± 0.7 cm). However, there were no differences in self-reported intake between these pairs (I). Analysis of obesity-discordant MZ twin pairs confirmed these results, as most pairs agreed that the heavier twins ate more (overall), snacked more and exercised less than the leaner co-twins. Eating behaviors such as eating too much, striving to be thin and body dissatisfaction were more frequently reported by obese co-twins than their lean counterparts. Total energy intake did not differ between the obese and lean co-twins and few differences were found in the 3-day food diary data. Underreporting of actual energy intake was significant for the obese (24.7% of TEE) but not for the lean co-twins (8.4% of TEE) (II). Obesity-discordant pairs differed significantly in their serum lipoprotein profiles, and the acquired accumulation of liver fat was particularly associated with increased concentrations of atherogenic lipids, including low-density lipoprotein (LDL-C) and Apolipoprotein B (ApoB), while physical activity was related to reduced concentrations of atherogenic lipids (IV). Omega-3 polyunsaturated fatty acid intake, as evaluated by self-reporting and serum measurements of DHA, was significantly related to a shift in the serum high-density lipoprotein (HDL) subclass distribution toward larger HDL particle size (III, V). In addition, serum DHA was positively associated with serum LDL particle diameter and negatively associated with triglyceride (TG) concentrations, medium and large very-low density lipoprotein (VLDL) particle concentrations and VLDL particle diameter. A high-fat, high-sucrose, low fiber dietary pattern (labeled junk food ) was positively associated with TG concentrations, a shift in the subclass distribution of VLDL toward larger particles and LDL toward smaller particles, and an increased concentration of small HDL particles in the serum (V). The associations were independent of adiposity and other lifestyle factors, and most were independent of the potential confounding effects of genotype and early environmental factors shared by twins. Conclusions: By using mutual responses of twin pairs, this study provides compelling evidence that acquired eating and physical patterns are important determinants of obesity, but they may be overlooked in population studies that use self-reported data due considerable misreporting of actual energy intakes and exercise behavior by obese subjects. Habitual physical activity and fish intake were related to a favorable serum lipoprotein profile, whereas a high-fat, high-sucrose, low-fiber dietary pattern and accumulation of liver fat associated with an unfavorable serum lipoprotein profile. These results emphasize a healthy lifestyle, in particular reduced portion sizes of energy-dense, nutrient poor foods and regular physical activity as the cornerstone of preventing obesity and lipid disturbances in young adults.