(1) The objectives of these studies were: (a) to determine the period during which fat cells were being formed in different fat depots in the pig; (b) to discover whether total fat cell number could be affected by overfeeding from weaning or later in life, and (c) to examine the domestic pig as a model for human obesity studies especially as to the influence of the energy intake at different stages in life on fat cell size and number. (2) Pigs were weaned at six weeks of age and allocated to one of four diets. Group H was fed to appetite on a high carbohydrate diet and became extremely obese. Group L was fed at half the energy intake of group H; these animals grew steadily but deposited little adipose tissue. Group L-H was fed as group L for 40 weeks and then as group H for the rest of the experiment, while group H-L was fed as group H for 40 weeks and then placed on a very restricted ("slimming") diet for the remainder of the experiment. (3) The volume of fat cells in group H was about four-fold greater than in group L at the end of the experiment. Cell volume increased rapidly when group L was fattened after 40 weeks and could be reduced significantly, albeit slowly, when group H was "slimmed". (4) There was a significant increase in the apparent number of fat cells in the depots of overfed pigs that did not diminish when the animals were "slimmed". However, in restricted pigs subsequently fattened, there was an apparent steady increase in fat cell number that eventually reached the same figure as in pigs overfed from weaning. (5) The standard methodology is inadequate to make broad generalizations about the influence of diet on cellularity. First, the development of fat cell size and number follows different time scales in different depots. It is essential to monitor several sites before attempting to estimate cell number. Secondly, since cells with diameters less than about 15 micrometer are not detected, "empty" cells, containing no fat, may be present from birth: subsequently they may be filled when energy intake is appropriate so that later they become visible, and are recorded as fat cells. In these cases there will be an "apparent" but not a real increase in cell number.