The problem of adaptive noisy clustering is investigated. Given a set of noisy observations \(Z_{i}=X_{i}+\epsilon _{i}\) , \(i=1, \ldots , n\) , the goal is to design clusters associated with the law of \(X_{i}\) ’s, with unknown density \(f\) with respect to the Lebesgue measure. Since we observe a corrupted sample, a direct approach as the popular \(k\) -means is not suitable in this case. In this paper, we propose a noisy \(k\) -means minimization, which is based on the \(k\) -means loss function and a deconvolution estimator of the density \(f\) . In particular, this approach suffers from the dependence on a bandwidth involved in the deconvolution kernel. Fast rates of convergence for the excess risk are proposed for a particular choice of the bandwidth, which depends on the smoothness of the density \(f\) . Then, we turn out into the main issue of this paper: the data-driven choice of the bandwidth. We state an adaptive upper bound using a modified version of Lespki’s method, called Empirical Risk Comparison, where empirical risks associated with different bandwidths are compared. Eventually, we illustrate that the selection rule can be used in many statistical problems of \(M\) -estimation where the empirical risk depends on a nuisance parameter.