The impact of macropore description on solute transport predictions in soils is not well understood. A 2-D Galerkin finite element model was used to compare different approaches for describing macropore flow in soil. The approaches were: a modification of the hydraulic conductivity function (Hydraulic function), the lumping of all macropores into one single straight macropore (Lumping), the use of an exchange factor between microporosities and macroporosities that occupy the same area (Dual porosity), and a detailed description of each macropore (Full description, base case). Simulated breakthrough curves were obtained with domains that contained one or more macropores of different shapes under both steady state and transient flow conditions. The Hydraulic function approach was not sensitive to macropore continuity and tortuosity. When the macropores were open at the soil surface and the solute was surface applied, the first three approaches underestimated both breakthrough curves and solute distribution in the profile compared to the Full description approach. When the solute was initially incorporated in the soil, the first three approaches overestimated the breakthrough curves compared to the Full description approach. The first three approaches also underestimated the heterogeneity of solute distribution in the profile compared to the Full description approach, mostly when the macropores were tortuous. The differences between predicted breakthrough curves with different approaches decreased with an increase in tortuosity and a decrease in surface continuity. To simplify macropore description, the Dual porosity approach was the better of the first three approaches for predicting breakthrough curves provided the exchange factor between macropores and matrix porosity was available.