This study has evaluated numerous analytical models for countercurrent capillary imbibition for use as matrix-fracture transfer functions in waterflood models for naturally fractured reservoirs. Relationships between the various models were developed and were used to reformulate the models into common terms so they could be compared. Laboratory imbibition data were obtained to validate the models. From this analysis, the strengths and weaknesses of each model were determined. An extended-square-root model was the best model for the early-time period when the finite size of the matrix block did not impact imbibition. An exponential model provided the best overall model for imbibition through the entire imbibition process, although it was not accurate for the very early times.