Field experiments were conducted to find the most productive and profitable row pattern in a groundnut/ pigeonpea combination under rainfed conditions. The treatments included two sole crops at their respective optimum population densities (333 000 plants ha-1 for sole groundnut and 60 000 plants ha -1 for sole pigeonpea) and five intercrop treatments involving groundnut/pigeonpea in 3:1, 4:1, 5:1, 6:1 and 8:1 ratios. The rows were sown 60 cm apart in sole pigeonpea and 30 cm in sole groundnut and all intercrop treatments. The intra-row spacing of both component crops in intercrop treatments was adjusted to maintain the population density equivalent to that of the sole crop optimum. Both 5:1 and 6:1 row arrangements were more productive in terms of yield advantage (LER of 1.49 for 5:1 and 1.48 for 6:1) and were also more profitable (5% monetary advantage for 5:1 and 3% monetary advantage for 6:1) than the 8:1 row arrangement (LER 1.41). Similar results were obtained in both years but the 6:1 row arrangement was more promising than the 5:1 row arrangement when the growing season was wet.