Two Lagrangian particle models for biomass growth in porous media were proposed. In the first model, a cellular automata approach was used to model interactions of biomass with a fluid. In the second model, pair-wise particle-particle interactions were used to simulate interactions within the biomass and both biomass-fluid and biomass-soil grain interactions. The biomass growth rate in both models was described by double Monod kinetics. For the set of parameters used in the simulations, both models produced qualitatively similar predictions: (1) biomass grows in the shape of bridges connecting soil grains and oriented in the direction of flow so as to minimize resistance to the fluid flow; (2) the solution containing electron donors and acceptors rapidly becomes depleted as it enters the fractured porous domain; and (3) the biomass growth occurs mainly at the entrance into the fracture. Quantitative predictions, such as total mass of microbes and spatial distribution of microbe concentration, were found to be strongly dependent on the type of model. The model that accounted for the detachment and attachment of the microbes predicted a higher ability of biomass to spread along preferential flow paths and seal off the porous matrix than the model with immobile biomass phase.