---

In order to investigate the roles of water harvesting patterns in water saving and increasing rain water productivity under micro catchment water harvesting systems, a field experiment was conducted in East Azarbaijan in the northwest of Iran during 1999-2006. The treatments included two MCWH patterns (small basins and semi-circular bunds), three catchment sizes (25m2 49m2 and 81m2), three runoff area treatments (Natural, cleared and smoothed, wetting and compacting) and two infiltration areas (Natural, soil mixed with polymer at a rate of 1kg/tree) for two new almond varieties. Polymer had no significant effect on water holding. Although the small basin with (9×9), runoff area compacted and use of polymer gave better results in survival, the growth and productivity of almond, the treatment with small basin (7×7) with compaction and without polymer can be recommended based on an economic analysis. Threshold runoffs were measured to be 2.5-3.5, 3.5-4.5 and 4.5-5.5 mm for natural, cleaned and smoothed, cleaned, wetted and compacted condite, respectively. Seventeen-year (1996-2013) average rainfall is equal to 202 mm, which shows that 84.4, 65.3, 51.3 and 36.4 of the total precipitation were more than 2, 4, 6 and 8 mm, respectively. Maximum rainfall runoff coefficient for threshold rain of 2, 4, 6 and 8 mm, was obtained to be 0.57, 0.35, 0.22 and 0.14, respectively. The design runoff coefficient for threshold rain of 2, 4, 6 and 8 mm, was obtained to be 0.28, 0.17, 0.11 and 0.07, respectively. Determination of threshold runoff has a major role in the success of water harvesting system design.