In recent years as costs for depositing dust and sludge generated in the iron and steel making industry have increased, due to lack of space and increasing environmental restrictions, the need to recycle these solid by- products to avoid depositing costs and to recover valuable metal fractions, is soon becoming a necessary reality. Cold bonded agglomeration is considered to be a well-suited alternative for recycling of steel industry by-product dust and sludge. The major objective of this work is to develop cold bonded pelletizing technology to increase the recycling of sludge and dust. Laboratory pelletizing tests were conducted based on a statistical procedure in order to later evaluate variables that effect cold strength, capacity and reduction of product pellets. The influence of BF flue dust, oily mill scale sludge, BOF fine and coarse sludge on the strength and capacity of cold bond pellets, using cement as binder, was studied experimentally. A related statistical procedure was used to conduct reduction experiments in inert gas over a temperature range of 20-1200 C. The results from cold strength and reduction tests have been evaluated using multivariate statistical analysis to model the experimental variables with given responses in order to help identify those variables that have most significance. Pellet blends with large fractions of particles in the size range of 10-40 microns promote good self-reduction while maintaining good cold strength. The fundamental reactions occurring during the heat treatment of cold bonded pellets have been studied. Blast furnace flue dust, which contains fractions of coal and coke particles, has been included in the cold bonded pellet blend as a source of solid reductant. Thermal analysis was performed on samples in inert atmosphere at a heating rate of 10 C/min in order to observe their high temperature properties, specifically, the mechanisms of self-reduction. The gases generated during thermal analysis were analyzed using a quadropole mass spectrometer. Furthermore, pellet samples were analyzed using X-ray diffraction, optical microscopy and scanning electron microscopy. Results demonstrate that the decomposition of hydrates and carbonates in cold bonded pellet samples contribute, as gaseous intermediates, to an earlier reduction of contained iron oxides. The gaseous intermediates are responsible for an initial gasification of carbon contained in blast furnace flue dust leading to low temperature iron oxide reduction. The step-wise reduction of iron oxides in the pellets at the given conditions begins at ~500 C and is nearly completed at 1200 C. The metallurgical characteristics of cold bonded pellets have been tested in additional laboratory tests i.e. isothermal reduction tests, blast furnace simulation and softening and melting tests. The test results indicate that cold bonded pellets can disintegrate during reduction at high temperature, are self-reducing to a high extent and, as a supplement to the normal ferrous burden, they show quite good softening and melting properties. Large scale trials with cold bonded pellets have been conducted in a commercial basic oxygen converter and a pilot scale blast furnace. Results from the basic oxygen furnace test show that charging of cold bonded pellets is feasible at levels up to 2.2% of the total charge weight. At these levels, cold bonded pellet addition resulted in no adverse disturbances to steel and slag chemistry and to the process in general. Results from the pilot scale blast furnace test show that the blast furnace operation was very stable during testing with 150 kg cold bonded pellets/tHM but the burden descent and gas distribution were disturbed during the periods with greater cold bonded pellet burden content. The rate of reducing agents was significantly decreased and slag amount was increased when cold bonded pellets were charged.