Kriging is a technique that has been used to interpolate a set of observed data points (x(,i), z(x(,i))); x(,i) (ELEM) DcR('p), i = 1, ..., n. The basic assumption is that the function z(x) is a realization of a stochastic process Z(x); x (ELEM) R('p). The relation between Kriging and BLUP is shown;We restrict our attention to covariance stationary stochastic processes, and we discuss the optimal allocation of the points x(,1), ..., x(,n) when p = 1 and D = 0,1;L-spline is another technique that is used for interpolation purposes. Although not always stated explicitly, there is a stochastic process that underlies an L-spline of interpolation such that Kriging and L-spline give the same interpolating function. The efficiency of the linear and cubic spline, with regard to Kriging, is evaluated for some covariance models that have been used in the geological area. The multiquadric method of interpolation is also compared with Kriging.