In this study, a field in situ inactivation experiment was carried out to control the confluence area sediment nutrient loading and fluxes using modified zeolite (MZ) in combination with poly aluminum chloride (PAC) and lanthanum-modified bentonite (LMB). The results indicated that PAC + MZ and LMB + MZ can reduce 76% and 75% of the P flux and 20% and 27% of the N flux, respectively. These results are based on a comparison with a control treatment over four months under the influence of external loading. However, their control efficiency on sediment nutrient fluxes decreased largely during the summertime algal blooming season. Both of the treatments lost their N control efficiency at this time. In contrast, LMB + MZ can still reduce 27% of the P flux compared to the control treatment. Surface sediment extractable ammonium increased substantially from the PAC + MZ and LMB + MZ treatments, which is 1.8 and 2.2 times more than the extractable ammonium in the control sediment after 210 days of remediation. The P fractionation analysis indicated that, in the PAC + MZ and LMB + MZ, both NaOH-rP and HCl-P increased greatly at a rate of 1.5 and 3.9 times, respectively, compared to the control sediment. PAC + MZ and LMB + MZ reduced the mobile P by 21% and 43%, respectively compared with the control sediment after 210 days of remediation. Bacteria richness and diversity in the PAC + MZ and LMB + MZ treatments had no obvious distinction when compared with the control treatment after 210 days of remediation but had a transient decrease in the LMB + MZ and recovered as it returned back to the same level found in control after 60 days. The results indicated that the control efficiency of nutrient fluxes in sediment might vary with types of inactivation agents and dosing methods and can be largely reduced under the influence of external loading and algal blooms.