Continuous decrease of the size of Thai rubber smallholdings for the last twenty years has led to general adoption of very intensive tapping systems by Thai rubber farmers. Beside 1/2 S d/2 7d/7, systems like ½ S d/1 2d/3, 1/3 S d/1 2d/3, 1/3 S d/1 3d/4 or even more intensive are common, resulting in over exploitation, high tapping panel dryness rates, short life-cycle of plantations and low productivity. Results of a three years experimentation on clone RRIM 600 in Chachoengsao Rubber Research Centre (RRIT-DOA) are presented. A new tapping system ("Double Cut Alternative" or DCA, 2x ½ S d/4 (t,t) 7d/7) is studied in comparison with ½ S d/2 7d/7 control. Principle of the DCA system is to optimise high tapping frequencies by alternate tapping of two cuts. The two cuts are opened on opposite panels (B0-1, B0-2) and maintained vertically distant by at least 75-80 cm, in order to minimise competition between their respective latex regeneration areas. The tree is tapped every 2 days, but each cut is tapped every 4 days, increasing regeneration time on each tapping cut (4 days instead of 2). After three years of tapping, DCA tapping system provides a significant increase in production (+25-30% for all yield parameters: kg/tree/year, kg/ha/year, g/tree/tapping and kg/tapper/day) compared to ½ S d/2 7d/7, together associated with improved latex physiological profiles (higher latex sucrose, inorganic phosphorus and reduced thiols). The DCA tapping system does not alter the competitive relationship between, trunk radial growth and rubber production. Compared to reduced tapping frequencies, which only increase output per tapper, DCA system increases both rubber yield and output per tapper in comparison with ½ S d/2 7d/7 conventional tapping system and therefore appears very promising for small-size rubber holdings, often forced to use higher tapping frequencies.