Understanding the spatial and temporal variations of infiltration in furrows is essential for the design and management of furrow irrigation systems. A key difficulty in quantifying the process is that infiltration depends on the depth of flow, which varies along a furrow and with time. An additional difficulty is that under many field conditions, a large fraction of the infiltrated water flows through cracks and/or macropores. This study examines the spatial and temporal variability of a furrow-irrigated field and evaluates a proposed semiphysical furrow infiltration model that accounts for flow-depth and macroporosity effects. Parameter estimation techniques were used to determine two parameters of the infiltration model, the hydraulic conductivity and the macroporosity term, in addition to the Manning roughness coefficient. The methodology was tested using published data from 30 furrow irrigation data sets collected in six furrows over five irrigation events. The evaluation revealed substantial variations in the final infiltrated volume among furrows and from one irrigation event to the next. Variability patterns differed markedly for infiltration measured during the advance phase in comparison with infiltration measured during the storage phase of the irrigations. Advance-phase infiltration varied systematically between irrigations for all furrows. Interfurrow inflow rate variability contributed to the variability of the infiltration during the postadvance phase, but not during the advance phase. Thus, cracks and/or macropores were an important contributor to the variability of infiltration during the advance phase. The analysis produced reasonable estimates of hydraulic conductivity relative to values reported in the literature. Hydraulic conductivity and post-advance infiltration volumes exhibited similar patterns of temporal variability. Hydraulic conductivity estimates were statistically correlated to the applied inflow rate. Although the reasons for this correlation are not clear, a possible explanation is that they are the result of systematic differences in the applied inflow rate among furrows. As with the advance-phase infiltration volumes, the estimated macroporosity term exhibited greater variation among irrigation events than among furrows during an event. The estimation procedure produced smaller differences between volume balance computed infiltration volumes and predicted values when using the semiphysical infiltration model than when using an empirical infiltration equation. Overall, the results show that the proposed furrow infiltration model represents the infiltration process adequately and that, at least for the studied data sets, the proposed estimation procedure yields a coherent set of infiltration and hydraulic resistance parameter values.