Endometriosis is a complex disease that affects 1 in 10 women worldwide. It occurs when tissue resembling the lining of the womb forms lesions in other parts of the body; predominantly within the pelvic cavity. The main symptoms of endometriosis are pain and infertility. Definitive diagnosis is through a surgical investigation which aims to identify lesions and assign a score based on the amount and location of lesions. This results in classification into four stages; minimal, mild, moderate and severe. This classification does not help to explain or predict disease diagnosis as some women with severe disease may not have any symptoms while others may have symptoms but no visible disease. Enhanced understanding of clinical features present in women with endometriosis may minimise the need for surgical investigation and improve treatment approaches. The aim of this study is to investigate if a patient’s clinical information could be used to predict whether endometriosis was present and the surgical stage of disease. This study included data from 251 women with suspected endometriosis. Clinical outcomes at surgery and other key clinical data were recorded. These included standard health data such as age and body mass index (BMI) as well as type and location of pain (e.g. lower back), fertility status and pregnancy history. Using this information, statistical analyses including univariate, multiple logistic and linear regression were performed to compare clinical data with surgical outcomes. Additionally, a more data-driven approach, known as conditional inference trees, was conducted to look at which combinations of clinical characteristics were associated with endometriosis diagnostic outcomes. It was found that women with moderate/severe stage endometriosis were more likely to be older (p<0.0001) and more likely to have infertility (p=0.0006). Pain was reported across the patient cohort but fewer women who were not diagnosed with endometriosis reported ovulation pain together with leg and lower back pain (p=0.04). Grouping key patient characteristics together in a multivariate linear regression model with disease score as the outcome showed that infertility (p=0.002), having no previous pregnancies (p=0.0002) and older age (p< 0.0001) were all associated a higher disease score indicating more severe disease. A multiple logistic regression model was fitted with an outcome of either no disease found or endometriosis. The odds of endometriosis being diagnosed were around 2 times higher in women who had been trying unsuccessfully to become pregnant for longer than 6 months and 3 to 4 times more likely in women with lower back pain. Analysis using conditional inference trees found that women who had pain on ovulation and current infertility (but no previous pregnancies) were more likely to be diagnosed with endometriosis (90%). Also, women with pain on ovulation, current infertility (but a previous pregnancy) as well as lower back pain were also more likely to be diagnosed with endometriosis (73%). Taken together, the data from this cohort of women suggest that older age, infertility, no previous pregnancies and pain on ovulation indicate a greater likelihood of being diagnosed with endometriosis or with a more severe surgical stage of the disease. These results provide evidence that using patient data could help to inform clinical decision-making and in the future aid identification of patients that may benefit most from a diagnostic laparoscopy.