Background Studies which have found that histological features of metastatic melanoma in SLN were predictive of survival, have differed considerably in their design and results. We investigated in detail the influence of SLN tumour, clinical and primary tumour parameters on clinical outcomes in a large cohort of patients treated in a single centre. Methods In SLN-positive melanoma patients, the association of clinical, primary tumour and SLN tumour features [maximum size (MaxSize), % cross-sectional area of SLN occupied by tumour (%CS), tumour-penetrative depth (TPD), intranodal location of tumour, extranodal spread (ENS) and perinodal lymphatic invasion (PLI)] with lymph node recurrence (LNR), disease-free (DFS), distant metastasis-free (DMFS) and melanoma-specific (MSS) survival was analysed. Results There were 409 SLN-positive patients. Age >50 years, presence of lymphovascular invasion in primary tumour and completion lymph node dissection (CLND) status were independent predictors of LNR. Primary tumour features (presence of ulceration and satellites) and presence of ENS in SLN were independent predictors of DFS, DMFS and MSS. In addition, poorer DFS was independently associated with primary tumour site (trunk vs limbs), SLN tumour features (MaxSize >2 mm, presence of PLI) and positive CLND; other factors independently predictive of DMFS were male sex, primary tumour features (absence of TILs) and SLN tumour MaxSize >10 mm; and age ≥50 years, presence of PLI in SLN were additional independent predictors of MSS. Conclusions The use of clinical, primary tumour and SLN tumour characteristics shown to be independent predictors of clinical outcomes in melanoma patients will assist in accurate prediction of prognosis and optimise clinical management.