During field application of biochar, the bulk density of tilled soil initially decreases and then increases over time, until reaching the initial level of compacted soil. This study evaluated the optimal biochar particle size for promotion of water infiltration and retention in a saline soil with various bulk densities after application. Corn straw biochar, pyrolyzed at 450 °C for 0.5 h, was prepared in different particle sizes (S1 ≤ 0.25 mm, S2 = 0.25–1 mm, and S3 = 1–2 mm) and separately mixed into the 0–30 cm soil layer at two rates (R1 = 10 g kg^-1 and R2 = 100 g kg^-1), with tilled (D1 = 1.1–1.43 g cm^-3) and compacted (D2 = 1.45 g cm^-3) bulk densities. Five models were applied to simulate water infiltration into biochar-amended soils. Compared with the non-biochar control, the S1 treatment increased cumulative water infiltration by 41% (bulk density = 1.26 g cm^-3) to 11% (bulk density = 1.45 g cm^-3). However, the effect of the S3 treatment on cumulative water infiltration shifted from positive (+19.3%) to negative (–22.4%) with increasing bulk density. The S2 treatment resulted in the highest water retention at the tilled bulk density, whereas a significant increase (12.7%) in water retention was observed in the S1 treatment at the compacted bulk density. The Kostiakov–Lewis, Kostiakov, and United States Department of Agriculture – Natural Resources Conservation Service models performed better than the Philip and Horton models to describe the relationship between cumulative water infiltration and infiltration time, except for the D2R2S1 treatment. This study provides evidence for amelioration of saline soil by straw biochar in the Yellow River Delta.