Survival for colorectal cancer patients in the UK lags behind other Western European countries. This thesis aims to explore the impact of delays in referral and missed opportunities in primary care on survival for colorectal cancer patients. For this purpose patients were identified from the Clinical Practice Research Datalink (CPRD), a national electronic primary care medical record database, with a diagnosis of colorectal cancer in the National Cancer Registry. The initial studies validate the use of CPRD to fulfil the aims of this thesis, by confirming patients included are representative of the wider national cancer population. In the latter chapters the association of delays in referral with survival are investigated using a retrospective cohort design, while adjusting for patient demographics and co-morbidities. Delays are defined as a referral after three months or after three or more consultations. In addition, missed opportunity for non-emergency diagnosis were identified for emergency patients with a previous consultation in primary care. Survival between the urgent and non-urgent referral pathways are also explored. This thesis demonstrates delays were associated with a worse survival and later stage for patients presenting with red-flag symptoms (as defined by NICE) but not non-specific symptoms. Presenting with non-specific symptoms was associated with a worse stage and emergency diagnosis independently. There was no difference in survival or stage for patients referred by the non-urgent pathway compared with the two-wait-week pathway. Identification of cancer patients in primary care is a difficult task. These findings suggest patients presenting with red-flag symptoms, and in particular rectal bleeding are presenting at an earlier stage providing an opportunity for a better prognosis if the referral is prompt. Alternative methods to differentiate cancer from benign patients need to be considered, as relying only on symptoms is insufficient.