The work was carried out in the Valle Cieneguillo Sur; Sector ATEA (Association of Agricultural Extension Technicians) to 9 km of the north of Piura, in a plantation of lemon tree of 7 years with distancing of 7x7 m2, lasted 22 months (January 2016 - November 2017), under sandy soil (92% sand) and 0.13% organic matter.The design Blocks Completely Random was used. The experimental unit was a tree in production.The factors under study were: 110 (N) 50 P2O5 120 K2O 25 Ca (D1), 210 (N) 100 P2O5 150 K2O 80 Ca (D2), 280 (N) 120 P2O5 180 K2O 120 Ca (D3), 50% Rice husk + 50% Organic matter (C1) and 75% Rice husk + 25% organic matter (C2); the combination determined the treatments: T1 (D1C1), T2 (D1C2), T3 (D2C1), T4 (D2C2), T5 (D3C1), T6 (D3C2) and T0 (without applying).With the results obtained, it was concluded: There are no statistical differences between the treatments on the number of first, second and third fruits.The fruit weight first, second and third, behaved statistically similar, however in the factory score, the control (481.67 g) behaved statistically equal to T5 (340.00 g) and is superior to the rest of the treatments.The T4 reported the best performance for the first lemon, with 7296.03 kg / ha. The T2 showed the highest performance for the second qualification with 2564.41 kg / ha. The yield (kg / ha) of third fruits was better with the treatments T5 and T6.