Bacterial cells are μm-sized ‘living machines’ capable of taking complex decisions. Despite being a unicellular, prokaryotic microorganism, the bacterium E. coli has an inherent mechanism to perceive the extracellular environment, take proper decisions depending on the nature of the stimulus and perfectly adapt to it. The underlying signal transduction mechanism of the organism is inherently noisy. In this talk, I will give an overview of a systematic and rigorous modelling approach, we developed to characterize the intrinsic noise in the biochemical sensory pathway of E. coli, and explore its effects. The model is used to predict quantities such as mean and fluctuations in the output, response of the pathway to external perturbations, chemotactic drift velocity and diffusion coefficient in an attractant gradient. Our analysis, analytical as well as computational, suggests that both methylation and signalling noise have significant impact on the chemotactic performance of the organism, in some parameter regimes.