Probability density function (PDF) methods have significant advantages in modeling turbulent combustion, in particular because the highly non-linear chemical source terms appear in closed form. The micromixing term in the PDF transport equations, representing diffusion in composition space, is however unclosed. Physically, micromixing occurs between regions of fluid having similar compositions. It is expected and generally shown that models that account for this localness of mixing perform better. In this work, a novel variation of the multiple mapping conditioning (MMC) mixing model is proposed. MMC makes use of reference variables to localise the mixing and was originally formulated for stochastic reference variables with standard Gaussian distributions. Here we follow the original interpretation but modify it by using a reference variable that has an evolving distribution, according to an Ornstein-Uhlenbeck process, with statistics that are similar to those of a physical scalar. In the present application of the model to nonpremixed combustion, mixture fraction is considered as the scalar; extensions of the model may be envisaged for other combustion modes by considering a different scalar variable. While the original version of MMC is mathematically elegant, it is conceptually complex and evaluation of model coefficients is difficult. The new model is both physically intuitive and its coefficients are easy to determine according to desired principles in canonical mixing situations. In Part I of this two-paper set, the model is studied analytically for homogeneous and mean scalar gradient flows without chemical reaction leading to an approach to set the model parameters that controls the unconditional scalar dissipation rate and delivers tunable localness. The behaviour of the model is also examined numerically in simple homogeneous mixing situations with and without chemical reaction. In Part II, the new MMC mixing model is implemented in the RANS context and validated against experimental data for the Sandia D-F flame series. (C) 2017 The Combustion Institute. Published by Elsevier Inc. All rights reserved.