To improve alfalfa silage quality and reduce additive cost, the potential of applying the fermented juice of epiphytic lactobacillus (FJEL) as an additive was evaluated. The experiment was designed as a 5 × 3 2-factor (wilt condition and additive) test. The second cut alfalfa at the budding stage was harvested, exposed to sunlight for 0 (20.04% DM; type 1), 5.2 (34.15% DM; type 2) or 8.5 h (54.58% DM; type 3). Alfalfa stuff were harvested in the next morning, exposed to sunlight and then natural rainfall for 1 h (33.20%DM; type 4) or 3 h (24.58% DM; type 5) before cutting. Alfalfa forage of each type was cut to 1–2 cm, then treated with distilled water (control), FJEL, or commercial lactobacillus product (CLP) at 10 ml/kg fresh weight (FW). The FJEL was made from alfalfa juice that was anaerobically stored for 2 d. The treated forages (approximately 500 g) were ensiled in vacuum-sealed polyethylene bags for 45 d, with 3 replicates for each treatment. The statistical significance of the effects of additive, wilt condition, and additive × wilt condition on silage quality were analyzed using a repeated measures one-way ANOVA. The application of FJEL decreased (P < 0.01) the pH value (4.44 vs. 4.93 or 4.66) and volatile fatty acid (VFA; 38.32 vs.49.55 or 44.82) content and increased (P < 0.01) the concentrations of lactic acid (LA; 68.99 vs.51.84 or 63.29) compared with those of the control and CLP treatment. However, the FJEL treatment had the lower (P < 0.01) aerobic stability (254 vs. 274 h), soluble fiber (148.89 vs. 154.85 g/kg DM), available NDF (145.39 vs. 154.25 g/kg DM), rapidly rumen degradable CP (59.91 vs. 62.33 g/kg DM) than CLP treatment. A higher pH and lower LA, sugar, starch and soluble fiber contents were found in the rain treatments compared with those of the no-rain treatments (P < 0.01). In conclusion, the application of FJEL could improve silage quality compared with the control; in addition, its effect as a fermentation stimulant may be comparable to or even better than CLP.