For different classes of deterministic and random sampling (t_k), we establish the asymptotic expressions for the bias and the variance of the estimate r_n(x) based on sampled data   for the regression function r(x) = E(Y_tX_t = x) of unbounded continuous-time processes   (not necessarily stationary). Under mild mixing conditions, we show that r_n (x) has exactly the same asymptotic quadratic error as in the i.i.d. case. In order to prove this result, we use some large deviations inequalities for mixing processes.