Equations for predicting the quantity of lean in the young pig were developed from measurements on 48 pigs (16.8 to 48.5 kg body weight) representing four crossbred mating types; maternal breed(s) × maternal breed(s), paternal breed(s) × maternal breed(s), paternal breed(s) × paternal breed(s) and porcine stress-susceptible × mixed breed. Within litter and sex, each of three pigs was assigned randomly to a light, intermediate or heavy slaughter weight group such that each mating type was represented by six barrows and six gilts from four different litters. Thirteen measurements were obtained for each pig; body weight, body length, body circumference, front leg circumference, front leg length, shoulder depth, shoulder width, ham width, head width, shoulder fat, last rib fat, last lumbar fat and loin muscle depth. Fat measurements and loin depth were taken ultrasonically. Pigs were slaughtered, chilled and manually separated into lean and fat, bone, skin and feet and tail components. Fat composition was determined by the Goldfisch method and also an x-ray absorption procedure. Pigs averaged 32.0 kg body weight and 12.5 kg of lean. Prediction equations were developed by maximum R² and stepwise regression procedures. The model that includes the 13 measured variables and average backfat produced a .97 R² when predicting kilograms of lean (LWTA). Body weight was the only significant variable. The one-variable model of body weight (P<.01) produced a .95 R² Kilograms of lean was estimated for an additional 24 pigs using the one-variable model of body weight and the two-variable model of body weight and shoulder width. The variable LWTA was found to be highly correlated with both estimates of LWTA (r = .97). Results suggest that LWTA can be predicted using a simple model that could be adapted to applied situations.