Since the early demonstration of 2D-electron gas [M. A. Khan et at., Appl. Phys. Lett. 60, 3027 (1992)] and a heterojunction field effect transistor (HFET) [M. Asif Khan et al., Appl. Phys. Lett. 63, 1214 (1993)] in III-N materials, rapid progress has been made to improve the DC and RF performance of GaN AlGaN based HFETs. Stable and impressive microwave powers as high as 4-8 W/mm have been reported for device operation frequencies from 10 to 35 GHz. The key reason for these high performance numbers is an extremely large sheet carrier densities (>1 x 10(13) cm(2)) that can be induced at the interfaces in III-N hetereojunction [A. Bykhovsk et al., J. Appl. Phys. 74, 6734 (1993); M. Asif Khan et al., Appl. Phys. Lett. 75, 2806 (1999)]. These are instrumental in screening the channel dislocations thereby retaining large room temperature carrier mobilities (>1500 cm(2)/Vs) and sheet resistance as low as 300 Omega/sq. These numbers and the high breakdown voltages of the large bandgap III-N material system thus enable rf-power approximately 5-10 times of that possible with GaAs and other competitor's technologies. We have recently introduced a unique pulsed atomic layer epitaxy approach to deposit AlN buffer layers and AlN/AlGaN superlattices [J. Zhang et al., Appl. Phys. Lett. 79, 925 (2001); J. P. Zhang et al., Appl. Phys. Lett. 80, 3542 (2002)] to manage strain and decrease the dislocation densities in high Al-content III-N layers. This has enabled us to significantly improve GaN/AlGaN hetereojunctions and the device isolation. The resulting low defect layers are not only key to improving the electronic but also deep ultraviolet light-emitting diode devices. For deep UV LED's they enabled us to obtain peak optical powers as high as 10 mW and 3 mW for wavelengths as short as 320 nm and 278 nm. Building on our past work [M. Asif Khan et al., Appl. Phys. Lett. 77, 1339 (2000); X. Hu et al., Appl. Phys. Lett. 79, 2832 (2001)] we have now deposited high quality SiO2/Si3N4 films Over AlGaN with low interface state densities. They have then been used to demonstrate III-N insulating gate transistors (MOSHFET (SiO2) and MISHFET (Si3N4) with gate leakage currents 4-6 order less than those for conventional GaN-AlGaN HFETs. The introduction of the thin insulator layers (less then 100 Angstrom) under the gate increases the threshold voltage by 2-3 V. In addition, it reduces the peak transconductance g(m). However the unity cut-off frequency, the gain and the rf-powers remain unaffected as the g(m)/C-gs (gate-source capacitance) ratio remains unchanged. In addition to managing the defects and gate leakage currents we have also employed InGaN channel double heterojunction structures (AlInGaN-InGaN-GaN) to confine the carriers thereby reducing the spillover into trappings states. These InGaN based MOS-DHFETs exhibited no current-collapse, extremely low gate leakage currents (<10(-10) A/mm) and 10-26 GHz rf-powers in excess of 6 W/mm. We have also demonstrated the scalability and stable operation of our new and innovative InGaN based insulating gate heterojunction field effect transistor approach. In this paper we will review the III-N heterojunction field-effect transistors progress and pioneering innovations including the excellent work from several research groups around the world. (C) 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.