FOUR EXPERIMENTAL DIETS, CONTAINING THE SAME BASAL COMPONENT BUT DIFFERENT AMOUNTS OF FIBER, WERE FED TO THIRTY, 24-GRAM CHANNEL CATFISH IN TWELVE, 20 SQUARE METER CONCRETE PONDS FOR 200 DAYS. THE BASAL COMPONENTS CONTAINED ALL KNOWN NUTRIENTS IN AMOUNTS ESSENTIAL FOR GROWTH, AND WERE DILUTED WITH PURIFIED WOOD CELLULOSE TO CONTAIN 8, 12, AND 20% FIBER. DISSOLVED OXYGEN AND AMMONIA LEVELS IN EACH POND WERE MEASURED. IN THE NEXT EXPERIMENT, DIGESTION COEFFICIENTS FOR PROTEIN, AVAILABLE CARBOHYDRATE, AND FAT WERE DETERMINED FOR THE DIETS CONTAINING THE FOUR LEVELS OF FIBER. THIS WAS CARRIED OUT IN RACEWAYS USING 400-500 GRAM CHANNEL CATFISH. FISH RECEIVING THE LEAST FIBROUS DIET GREW SIGNIFICANTLY MORE (P>0.05) THAN THOSE RECEIVING GREATER AMOUNTS OF FIBER. DURING THE LAST 30 TO 60 DAYS OF THE FEEDING PERIOD, DISSOLVED OXYGEN AVERAGES WERE REDUCED FOR THOSE RECEIVING THE HIGHER FIBER DIETS. DIGESTION COEFFICIENTS FOR THE THREE NUTRIENTS WERE ESSENTIALLY UNAFFECTED BY THE FIBER CONTENT OF THE DIETS. THE DATA FROM THIS STUDY INDICATE THAT FIBER DOES NOT APPEAR TO BE A NUTRITIONALLY IMPORTANT COMPONENT OF PRACTICAL FEEDS FOR CHANNEL CATFISH. THE SLIGHTLY HIGHER WEIGHT GAINS BY THE FISH FED THE DIETS CONTAINING THE LOWEST FIBER LEVEL CANNOT BE RELATED TO NUTRIENT DIGESTIBILITY. THE CAUSE FOR THE DIFFERENCE IN GROWTH RESPONSE MAY BE DUE TO FISH NOT CONSUMING ALL OF THE HIGHER DIETS OR TO STRESS OF POORER WATER QUALITY DURING THE LATTER PHASE OF THE EXPERIMENT.