Aerobic granular sludge (AGS) in sequencing batch reactors (SBR) has recently made its proofs as full-scale technology for the biological treatment of urban wastewater. In contrast to conventional flocculent activated sludge, microorganisms aggregate to dense granular biofilm due to shear stress from aeration. Hence, AGS has much better settling characteristics and allows the construction of more compact wastewater treatment plants (WWTP) without secondary clarifiers. Due to diffusion limitations of oxygen through the biofilm, aerobic and anaerobic/anoxic zones coexist in AGS. Hence, AGS-SBR has the potential to treat organic matter (COD), nitrogen (N), and phosphorous (P) simultaneously in one single reactor. The focus of the present PhD thesis was on the startup of AGS-SBR and the optimization of biological nutrient removal (BNR). For the investigation of the optimal startup conditions, a study testing seven parameters in parallel was conducted. The main conditions identified for a rapid startup of AGS-SBRs with good nutrient removal performances were (i) the alternation of high and low dissolved oxygen (DO) phases during aeration, (ii) a settling strategy avoiding too high biomass washout, (iii) the adaptation of the pollution load in the early stage of the startup in order to ensure that all soluble COD was consumed before the aeration phase, (iv) higher temperature (20°C), and (v) a neutral pH. Under such conditions it took less than 30 days to produce granular sludge with high removal performances for COD, N and P. A control run of the best startup strategy led to very similar results, proving the reproducibility of the experimental approach. This control run was operated for 80 days without any problems concerning the stability of the granular sludge or the nutrient removal performances. Concerning the bacterial community composition during the startup phase, a general shift of the predominant populations from Intrasporangiaceae and Sphingobacteriales to Dechloromonas and Zoogloea was observed. This shift was mainly due to general conditions with lab-scale AGS-SBR, rather than the specific operation parameters tested in the different experimental runs. However, it has been observed that polyphosphate-accumulating organisms (PAO) and glycogen-accumulating organisms (GAO) related populations were favored by the adaptation of the pollution load in the early stage of the startup in order to ensure that all soluble COD was consumed before the aeration phase. Besides the pollution load, the temperature and the pH had a significant impact on the global bacterial community structure. The predominant PAO were presumably Dechloromonas. In order to optimize BNR by AGS-SBR, different aeration strategies were tested. It has been concluded that the N-removal efficiency of COD-limited systems can be considerably enhanced with improved aeration strategies. Strategies promoting alternating nitrification and denitrification (AND) were significantly more efficient than simultaneous nitrification and denitrification (SND) strategies. The introduction of low DO phases or even anoxic phases in an early stage of the total aeration period probably enhanced denitrifying P-removal and led to COD savings. Intermittent aeration, which is a realistic AND strategy for full scale applications, led to the highest N-removal efficiency. The short mixing times implemented with this strategy were not problematic for the stability of the granules. Finally, to further optimize N-removal in COD-limited systems, the potential of aeration control for the achievement of N-removal over nitrite was investigated. It was shown that aeration phase length control combined with intermittent aeration or alternating high-low DO, is an efficient way to achieve N-removal over nitrite. N-removal efficiencies of up to 95% were achieved with this way of reactor operation. At 20°C, N-removal over nitrite was achieved within 20 –– 60 days and it was possible to switch from N-removal over nitrite to N-removal over nitrate and back again. At 15°C, the nitrite-oxidizing bacteria population could be reduced, but nitrite oxidation could not be completely repressed. However, the combination of aeration phase length control and high-low DO was successful to maintain the nitrite pathway at 15°C where the maximum growth rate of nitrite-oxidizing bacteria is clearly higher than the one of ammonium-oxidizing bacteria. In conclusion, this thesis showed the potential of the optimization of operation conditions for the startup of AGS-SBR and BNR. An efficient strategy for the startup with flocculent inoculum sludge has been developed, maintaining high BNR. Moreover, it has been shown that N-removal under COD-limited conditions can be improved by aeration control, either by enhancing denitrifying P-removal or achieving N-removal over nitrite.