Certain techniques of measuring differences in groups of experimental fish such as might be used in testing hatchery diets or environmental factors have been developed. In an experiment of 2 monthsˈ duration, nine similar groups of fish were weighed as groups at weekly intervals, and measured individually on four occasions to determine the degree of uniformity in growth attainable under experimental conditions. The standard deviation of length for each group maintained a fairly constant relationship to the other groups. Rate of growth was rapid. The average weight in grams at any time equalled (2.9594) (1.2378)t where the exponent t is time in weeks. The 20-ingredient diet is tabulated with percentages; an over-all food conversion factor of 3 to 1 was obtained. Substitution of halibut viscera for 20 percent of the regular diet in four of the groups reduced the rate of growth significantly and was discontinued after 3 weeks. A group-photographic method of recording size was tested and found inferior to manual measuring of length for general purposes, although it had advantages in speed of obtaining a permanent record of size and general appearance. Tests of three of the length-frequency distributions showed them to be approximately normal.