<p>An unbiased estimate of Gutenberg-Richter’s <i>b</i>-value and its uncertainty is the most important step for performing a realistic seismic hazard assessment. Besides this, the <i>b</i>-value gives important information on the process of earthquake generation, earthquake forecasts and the physics of earth crusts and therefore, an unbiased estimate of <i>b</i>-value is a necessity. An unbiased estimate of <i>b</i>-value to an acceptable degree of accuracy depend on the quality of catalogue, amount of data available for a region under study, discretization of the continuous magnitudes and imposition of an upper bound magnitude. This paper aims to compute an unbiased <i>b</i>-value by addressing various important issues that generally find little space in literatures. The results are presented for India subcontinent by dividing it into different provinces, characterized by distinct seismotectonic and geotechnical nature. Despite the fact that limitations and uncertainties are inherent to any earthquake catalogue, the effect of biases arising out of discretization of continuous magnitudes and imposition of an upper bound magnitude are elaborated with candid mathematical expressions while computing an unbiased estimate of <i>b</i>-value. The paper also presents the mapping of <i>b</i>-values, grouped into different intervals, across India and the change in the mean annual rates of exceedance for large magnitudes due to the truncated Gutenberg-Richter law for all the seismotectonic provinces of India. The results will be useful in future research for carrying out any seismicity-based studies and especially probabilistic seismic hazard assessment by zoneless approach, which accounts for the non-uniform probability distribution of earthquakes inside a seismotectonic province.</p>