An alteration in the state of climate that lasts for a longer duration of time, usually decades or more, which can be detected by variations in mean and/or variability of its characteristics is known as climate change (CC). The “detection” of CC is the process of showcasing that climate or a system impacted by climate has altered in some definite statistical manner, without giving a cause for that change, whilst “attribution” is the method of examining the relative contribution of various causal aspects to a change or event with determination of statistical confidence, as per the good practice guidance paper of IPCC on detection and attribution (D&A) of CC. It is necessary to understand the CC occurring all over the world and the reason behind it. Therefore, this study has reviewed various methods and advances in methodologies used for CC D&A during the past few decades. For the trend analysis, both parametric and non-parametric methods are used in most of the reviewed studies. Fingerprinting, optimal fingerprinting and artificial neural network (ANN) methods are the three main attribution techniques found in the literature. The fingerprinting (FP) method has the drawback of reduced signal-to-noise ratio. Optimal fingerprinting (OFP) uses multivariate regression which has four basic assumptions, and it is difficult to satisfy these assumptions. Therefore, the result of multivariate regression may not be reliable. The climate system is considered to be a nonlinear system and ANN models are non-linear, so ANN models are useful for attribution of CC.