In southern Thailand, rubber smallholders normally use very intensive tapping systems. This may result in overexploitation, high tapping panel dryness rates and a short production lifespan with low productivity. A "double cut alternative" (DCA) tapping system was tested in Songkhla province, a traditional rubber production area. The aim of the DCA system is to optimize high tapping frequencies by splitting tapping between two different cuts tapped alternately to increase the latex regeneration time in the bark. An on-station trial was started in 2007 at the Thepa Research Station, Songkhla province. The experiment used a "one tree plot design" comprising 20 replicates per treatment with four different treatments in a plot planted in 1999 with the rubber clone RRIM 600. During the fi rst three years of tapping, the DCA treatment T2 led to a 22% increase in yield (kg tree-1) compared with its control (T1). The DCA treatment T4 led to a 16% increase in yield compared with its control (T3). Output per tapping (g tree-1 tapping-1) in the DCA treatments also increased. There was no adverse impact on radial trunk growth, but the two DCA tapping systems led to a 13-19% increase in bark consumption. The results of the on-station trials using the DCA tapping systems were promising, but on-farm trials with farmers in different areas in southern Thailand are needed before this technique can be recommended at a larger scale.