Deterioration in soil physical quality under intensive tillage practices is a norm rather than an exception. The objectives of this study were to (i) evaluate total porosity (TP) as an indicator parameter to assess the impact of banana cropping on compaction and infiltration in soils, and (ii) assess the effectiveness of different interrow grass-covers in minimising the deteriorating trends. Depth-incremented TP under banana rows and interrows from 4 sites, corresponding forest sites, and from 3 interrow grass-cover treatments were computed from bulk density measurements. The TP results show the compacted depth ranged from 0.35 to 0.45 m in banana rows and from 0.35 to 1.0 m in the interrows. The TP in 0.10 m depth increments decreased in the order: forest > rows > interrows, and was positively correlated with soil organic C (OC) and negatively with wheel traffic stress (WTS). The multiple regression analysis showed that 77% of the variability in TP was accounted for by clay + silt, OC, and WTS. We show that a threshold compaction index (DI_t) of 0.81–0.83 can be estimated from TP regardless of the soil type. Depending on the soil type and the cultural practices followed, infiltration decreased from 0.75 mm/s in rainforest to 0.23 mm/s under banana in 1 soil type compared with 2.55 mm/s in forest and 0.85 mm/s under banana in another. After 18 months of interrow grass-covers we found the deterioration in TP was minimum under the indigenous grass-cover but not under the 2 improved species. We conclude the interrow grass-covers were effective in minimising WTS associated compaction and reduction in infiltration.