Intramuscular fat content is generally associated with improved sensory quality and better acceptability of fresh pork. However, conclusive evidence is still lacking for the biological mechanisms underlying i.m. fat content variability in pigs. The current study aimed to determine whether variations in i.m. fat content of longissimus muscle are related to i.m. adipocyte cellularity, lipid metabolism, or contractile properties of the whole muscle. To this end, crossbred (Large White x Duroc) pigs exhibiting either a high (2.82 +/- 0.38%, HF) or a low (1.15 +/- 0.14%, LF) lipid content in LM biopsies at 70 kg of BW were further studied at 107 +/- 7 kg of BW. Animals grew at the same rate, but HF pigs at slaughter presented fatter carcasses than LF pigs (P = 0.04). The differences in i.m. fat content between the 2 groups were mostly explained by variation in i.m. adipocyte number (+127% in HF compared with LF groups, P = 0.005). Less difference (+13% in HF compared with LF groups, P = 0.057) was noted in adipocyte diameter, and no significant variation was detected in whole-muscle lipogenic enzyme activities (acetyl-CoA carboxylase, P = 0.9; malic enzyme, P = 0.35; glucose-6-phosphate dehydrogenase, P = 0.75), mRNA levels of sterol-regulatory element binding protein-1 (P = 0.6), or diacylglycerol acyltransferase 1 (P = 0.6). Adipocyte fatty acid binding protein (FABP)-4 protein content in whole LM was 2-fold greater in HF pigs than in LF pigs (P = 0.05), and positive correlation coefficients were found between the FABP-4 protein level and adipocyte number (R2 = 0.47, P = 0.02) and lipid content (R2 = 0.58, P = 0.004). Conversely, there was no difference between groups relative to FABP-3 mRNA (P = 0.46) or protein (P = 0.56) levels, oxidative enzymatic activities (citrate synthase, P = 0.9; beta-hydroxyacyl-CoA dehydrogenase, P = 0.7), mitochondrial (P = 0.5) and peroxisomal (P = 0.12) oxidation rates of oleate, mRNA levels of genes involved in fatty acid oxidation (carnitine-palmitoyl-transferase 1, P = 0.98; peroxisome proliferator-activated receptor delta, P = 0.73) or energy expenditure (uncoupling protein 2, P = 0.92; uncoupling protein 3, P = 0.84), or myosin heavy-chain mRNA proportions (P > 0.49). The current study suggests that FABP-4 protein content may be a valuable marker of lipid accretion in LM and that i.m. fat content and myofiber type composition can be manipulated independently.