Although geotextile mats have considerable potential for soil conservation, field studies on their impacts on soil particle size distribution and nutrient conservation are scant. Hence, field experiments were conducted at Hilton, east Shropshire, U.K. (52°33′5.7″N, 2°19′18.3″W) during 2007–2009, to study the impacts of palm-mat geotextiles on topsoil (0–5 cm) particle size distribution and changes in selected nutrients (total P, K, Ca, Mg, S, Zn, Cu, Fe, Mn, Mo and Cl). Geotextile-mats constructed from Borassus aethiopum (Borassus palm of West Africa) and Mauritia flexuosa (Buriti palm of South America) leaves are termed Borassus mats and Buriti mats, respectively. Ten runoff plots (10 × 1 m on a 15° slope) were established, with duplicate treatments. The treatments were: (i) bare soil; (ii) permanent grass; (iii) bare soil with 1 m Borassus-mat buffer zones (area coverage ~ 10%) at the lower end of the plots; (iv) bare soil with 1 m Buriti-mat buffer zones (area coverage ~ 10%) at the lower end of the plots; and (v) completely-covered with Borassus mats. Initial and final topsoil samples were collected and analysed for sand (fine, medium and coarse), silt (fine, medium and coarse), clay and selected nutrients. Results revealed that soil silt and clay contents significantly ( P < 0.05) decreased in the Borassus completely-covered plots with a concomitant increase in proportion of sand content after ~ 2 years. Both Borassus and Buriti buffer strip plots also had lower soil clay contents, indicating that palm-mat geotextile cover significantly affected particle size distribution, even after only 2 years. Buriti mat-cover within the buffer strip plots resulted in notably higher decrease in mean ( n = 2) clay content, with a concomitant increase in mean ( n = 2) sand contents compared with mean ( n = 18) clay and sand contents of bare soils within the same plots. Data indicate that total P and total K contents with Borassus completely-covered plots significantly ( P < 0.01) increased and total Ca content decreased after ~ 2 years. However, except grassed plots, all plots had significant ( P < 0.05) increases in total P concentrations. Borassus buffer strip plots also had significant ( P < 0.05) decreases in total Ca contents over the initial contents. The increases in total P and K contents within Borassus complete-cover plots were significantly ( P < 0.05) higher than both grassed and Buriti buffer strip plots. Borassus buffer strip plots significantly ( P < 0.05) decreased total Ca content compared to bare and grassed plots. Treatments had no significant ( P > 0.05) effects on changes in other selected nutrients (total Mg, S, Zn, Cu, Fe, Mn, Mo and Cl) concentrations. In summary, it is evident that despite significant ( P < 0.05) decreases in soil clay and total Ca contents, use of Borassus mats as buffer strips was very successful in conserving or improving other selected soil properties. ► Impacts of geotextiles on soil texture and nutrient retention were studied ► Buffer strip plots had lower soil clay contents than bare plots ► Borassus complete-covered plots had less clay and silt contents than initial soil ► Total P content increased in the Borassus complete-covered plots over initial soil ► Mat cover had no effect on changes in total Mg, S, Zn, Cu, Fe, Mn, Mo and Cl