We examined crop yields along a wide environmental gradient and spatial dynamics in soil organic matter in response to conservation farming (CF) in Zambia. Maize yields from farmer managed CF and traditional farming (TF) were not significantly different with over 280 on-farm trials with varying soil properties, management practices and environmental covariates. Principle component analysis (PCA) identified inappropriate management practices (planting and insufficient weeding), of which 25% of total variability were major factors restricting CF yields. TF yields were limited by both amount and types of inputs that explained 26 % of total variability. With addition of different organic and inorganic amendments, average CF yield ranged from 1 to 4 t ha-1, with highest in wetter region (3.4±7.9 t ha-1) and lowest (2.1±6.8 t ha-1) in degraded plateaus. Combined additions of inorganic fertilizer (NP-K at 200-100-100 kg ha-1) with biochar and manure achieved the highest effect in degraded plateau with yield increase of 320% and 300% respectively as compared to organic matter additions of manure (46%) and gliricidia (24%) in the same region. PCA established pre-existing soil fertility is the major factor in all sites for improved yields and nutrient uptake with organic additions (P<0.05). Compost additions (P=0.001), and manure with or without inorganic fertilizer additions (P=0.02) led to greater yields with finer soil texture but not with biochar additions. iii Additions of biochar with inorganic fertilizer in wetter region enhanced maize Ca uptake (P=0.03) at lower pH (P=0.005) and higher rainfall (P=0.05). Total soil organic C (SOC) and N contents were initially 8% and 12% greater in planting basins than in rows over 10-year chronosequence under CF. Both SOC and N contents increased to a greater extent in basins than in rows with increasing years indicating greater SOM accrual. Mineralization of C per unit SOC in basins (R2=0.83) increased with years under CF indicating an accumulation of more labile SOC, whereas no changes were observed in rows. Potential mineralized N (PMN) increased in both basins (R2=0.60) and rows (R2=0.79) although more rapidly in basins than in rows. Greater stability of SOC was observed in areas receiving crop residues only. iv