Background Currently there is no universally agreed schema for predicting ocular morbidity in facial nerve palsy. The House Brackmann Scale has limitations in assessing ocular morbidity from facial nerve palsy. Our aim was to create a scoring system to help quantify ocular morbidity to aid in decision making regarding the need for corneal protective oculoplastic surgery. Methods We conducted a large cohort study observing 606 patients attending the specialist facial palsy clinic in Manchester UK between March 2002 and October 2017. Retrospective multivariate analysis identified clinical predictors for the 316 patients that required oculoplastic surgery. β coefficients generated in the multivariate analysis helped formulate a new facial nerve palsy scoring instrument to predict the need for corneal protective oculoplastic surgery. Results The House Brackmann Scale, corneal lagophthalmos and loss of corneal sensation proved clinically significant predictors for requiring corneal protective oculoplastic surgery. The scoring system derived from these factors provided an accurate and repeatable prediction tool demonstrated by validation studies on our patient population. The area under the ROC curve for the multivariate prediction model was 0.769 (0.726, 0.811). A score of 5 points out of a possible 8 was the best cut off score to recommend oculoplastic surgery, giving a sensitivity of 0.750 and a specificity of 0.671. Conclusion This study demonstrates that corneal lagophthalmos, corneal sensation and the House Brackmann Scale are important in predicting the need for corneal protective oculoplastic surgery. Our scoring tool is an important clinical decision tool for ophthalmic and ENT colleagues.