Abstract: Two field experiments were conducted to explore the influence of grain amaranth (Amaranthus hypochondriacus) and tobacco intercropping on tobacco growth and potassium uptake during 2014 and 2015. The experiment during 2014 included three treatments, (a) planting tobacco only; (b) tobacco and amaranth intercropping, amaranth was cut and burried into soil after 45 day growth; (c) tobacco and amaranth intercropping, but did not cut and burry. There were 4 treatments during 2015, (d) planting tobacco only; The other 3 treatments were all tobacco and amaranth intercropping, (e) one row of tobacco with 2 rows of amaranth; (f) one row of tobacco with 4 rows of amaranth; (g) one row of tobacco with 6 rows of amaranth. We analyzed tobacco dry matter accumulation, tobacco potassium concentration, potassium uptake, soil available potassium and soil slow-release potassium in the three intercropping systems. The results showed that, compared with treatment (a), amaranth intercropping without returning to the field (treatment c) increased dry matter accumulation in tobacco roots. Potassium concentration showed no significant changes, while potassium content increased by 70%. Dry matter accumulation reduced by 11% and 14% respectively in stem and leaves, while potassium contents were 0.11% and 0.12% lower and potassium concentration 16% and 12% lower. For the treatments with amaranth returning to the field, the highest potassium concentration in leaves was recorded. The different intercropping systems showed different amounts of dry matter accumulation, potassium content and potassium concentration. Dry matter accumulation and potassium contents in root, stem, upper leaves and middle leaves increased in the 2 rows (e) and 4 rows (f) treatments and decreased in the 6 rows (g) treatment. Dry matter accumulation in the upper and middle leaves was lower in treatment (e) than (f), but the former's potassium concentration and potassium content were higher. The concentration of soil available potassium in the tobacco rows increased in the intercropping systems, but the concentration of soil slow-release potassium did not change. The intercropping systems without grain amaranth returning showed lower soil available potassium concentrations than those with amaranth returning. Among the treatments in the amaranth returning system, the growth in soil available potassium was largest in treatment (g), which increased by 91% and 35% when returned after 20 and 60 d respectively. Soil available potassium increased by 54% and 23% in (f) over these periods. The analysis of apparent potassium flow showed different results for the various amaranth returning systems. About 42%-52% of potassium was uptaken by the tobacco in treatments (e) and (f), while only some 17% was absorbed in treatment (g). This study suggests that the most favorable planting pattern is the amaranth-tobacco intercropping system with 2 rows of amaranths and returning amaranth to the field after 45-55 d growth.