The feasibility of future electron-positron colliders operating at energies >1 TeV will depend on both operating efficiency and cost of the microwave amplifiers that can be developed to drive the collider. To zeroth order, the required number of amplifiers depends inversely on the parameter A = P-p tau(p)/lambda(2) where lambda is the wavelength, and P-p and tau(p) are, respectively, the power and duration of the amplifier output pulses. Thus, one goal of amplifier research is to maximize A while keeping other parameter values within bounds so as not to excessively increase the cost of either the individual amplifier system or the collider structure (e.g., amplifier voltage V less than or similar to 500 kV, wavelength lambda greater than or similar to 1 cm). operating within these bounds, gyro-amplifiers have demonstrated values of A = 11 x 10(4) W s/m(2), which compares favorably with the best values of A achieved by klystrons. The gyro-amplifier program which led to this accomplishment is reviewed, Some 20 different gyro-amplifier configurations have been examined on our 450-kV gyro-amplifier test facility during the past several gears. These tubes fall into five major categories: first-, second-, and third-harmonic gyroklystrons, as well as first- and second-harmonic gyrotwystrons. Peak powers in excess of 30 MW with pulse duration of 0.8 mu s have been achieved at 19.76 GHz in the TE(02) mode via a two-cavity second-harmonic gyroklystron with a first-harmonic drive cavity. The peak efficiency and gain were 28% and 27 dB, respectively. At present, there is ongoing construction of a new three-cavity second-harmonic coaxial gyroklystron, driven by a 500-kV 720-A beam, which is predicted to have an output power wed above 100 MW at 7.136 GHz with an intrinsic efficiency in excess of 40%. With the use of a depressed collector, achievable overall amplifier efficiency, which is of very great importance in the collider application, could be in the range of 50-65%.